MEDITERRANEAN ARCHAEOLOGY & ARCHAEOMETRY Editorial Board Archaeology Dr. Vincenzo Bellelli (Consiglio Nazionale delle Ricerche, Italy) Prof. Anna Belfer Cohen (University of Tel Aviv) Dr. David Blackman (Oxford) Dr. Mary Blomberg (Uppsala) Prof. Eric H. Cline (The George Washington University) Prof. John Coleman (Cornell University) Dr. Massimo Cultraro (Inst. per i Beni Archeol. e Monument.) Dr. Jack L. Davis (University of Cincinnati) Prof. Herald Hauptmann (Heidelberg) Prof. Vladimir I. Ionesov (Samara State Academy of Culture and Arts) Dr Omar Kareem (Cairo University) Prof. Bernard Knapp (University of Glasgow) Prof. Janusz Kozlowski (University of Crakow) Dr. Nina Kyparissi-Apostolika (Greek Ministry of Culture) Prof. Irene S Lemos (University of Oxford) Prof. Mehmet Ozdogan (University of Istanbul) Prof. Luiz Oosterbeek (do Instituto Politécnico de Tomar) Dr. Simon Stoddart (University of Cambridge) Mrs. Marta Santos Retolaza (Museu d'Arqueologia de Catalunya-Empúries) Dr. Chris Stevenson (Virginia Commonwealth Univ.) Prof. Emer. Petros Themelis (University of Crete) Prof. Rene Treuil (University of Paris X) Prof. Assaf Yasur-Landau (University of Tel Aviv) Prof. Willeke Wendrich (UCLA, Cotsen Institute of Archaeology) Prof. El-Sayed Mahfouz (Assiut University) Prof. Christophe Morhange (Université d'Aix-Marseille) Archaeometry Dr. Grzegorz Adamiec (Silesian University of Technology) Prof. Juan Barcelo (Universitat Autonoma de Barcelona) Dr Cathy Batt (University of Bradford) Dr. Michael Baxter (University of Nottingham) Dr. Jaume Buxeda i Carrigos (Univ. of Barcelona) Dr. Joachim Burger (Mainz University) Prof. Maria Perla Colombini (Univerrsita di Pisa) Dr. Paul Craddock (The British Museum) Dr. Martin P. Evinson (University of Sheffield) Prof. Mauricio Forte (University of California) Dr. Michael Glascock (University of Missouri) Prof. Philippe Lanos (Université de Rennes 1) Prof. Giulio Magli (Politecnico di Milano) Prof. Rocco Mazzeo (University of Bologna) Dr. Andrew Murray (University of Aarhus) Dr. Anna Pazdur (Silesian University of Technology) Prof. Emer. Vassilis Perdikatsis (Technical Univ. of Crete) Dr. Phil Potts (Open University, UK) Dr. Paula J. Reimer (Lawrence Livermore) Prof. Ashok Singhvi (PRL Ahmedabad) Prof Rob Sternberg (Franklin & Marshall College) Prof. Gregory Tsokas (Aristotle Univ. of Thessaloniki) Robert H. Tykot (University of South Florida) Prof. Andreas Vött (Johannes Gutenberg-Universität Mainz) Dr. Ian Whitbread (University of Leicester) Dr Kevin Walsh (The University of York) Dr Steve Wiener (Kimmel Center for Archaeological Science) Editor-in-Chief Professor Ioannis Liritzis University of the Aegean Department of Mediterranean Studies Laboratory of Archaeometry Rhodes 85100, Greece Tel & Fax: +30 22410-99320 & +30 22410 99385-6 e-mail: liritzis@rhodes.aegean.gr Editorial Office Technical Support & Handling Th. Tsigaros (M.Sc., University of Aegean) Ass.Prof. S.Vosynakis (University of the Aegean) Editors Archaeology Prof. Anagnostis Agelarakis (Adelphi University) e-mail: agelarak@adelphi.edu Dr. Ann Brysbaert (University of Leicester) e-mail: anb11@le.ac.uk Prof. Zeidan Kafafi (Yarmouk University) e-mail: zeidank@yahoo.com Dr. Ioannis Papadatos (University of Athens) e-mail: ypapadatos@yahoo.gr Archaeometry Julian Henderson (University of Nottingham) e-mail: Julian.Henderson@nottingham.ac.uk Ioanna Kakoulli (University of California, Los Angeles) e-mail: kakoulli@ucla.edu Prof. Marco Martini (University of Milano - Bicocca) e-mail: m.martini@unimib.it Asoc.Prof. Nikos Zacharias (Univ. of Peloponnese) e-mail: zacharias@uop.gr No part of this publication may be reproduced partly or wholly, in summary or paraphrased, stored in a retrieval system or transmitted, in any form, or by any means, electronic, mechanical, photocopying, recording or other without the prior permission of the publishers. Law 2121/1993 and the rules of International Law applying in Greece. Copyright © 2014 MAA International Journal All rights reserved ΜΑΑ is indexed and abstracted in ARTS AND HUMANITIES CITATION INDEX (Thomsons Reuters), SCOPUS (Elsevier), EBSCO, ULRICH’S, NASA-ADS, GOOGLE SCHOLAR Publication and Subscription Rates MAA is published twice per year in paper and in electronic edition. Interested readers should contact the editor-in-chief via e-mail. Subscription is as follows: 100€ (Libraries), 80€ (Individuals), 50€ (Students). Correspondence and Inquiries All correspondence and inquiries regarding editorial matters see: www.maajournal.com MEDITERRANEAN ARCHAEOLOGY & ARCHAEOMETRY www.maajournal.com AIMS AND SCOPES The Mediterranean Archaeology & Archaeometry (MAA) is an established interdisciplinary International Journal, running steadily since 2001, issued by The University of the Aegean, Department of Mediterranean Studies, Rhodes, Greece. It focuses in the Mediterranean region and on matters referred to interactions of Mediterranean with neighboring areas, but presents an international forum of research, innovations, discoveries, applications and meetings, concerning the modern approaches to the study of human past. It covers the following interdisciplinary topics: ■ theoretical & experimental archaeology, ■ environmental archaeology ■ ethnoarchaeology ■ completed excavation reports, ■ palaeolithic, prehistoric, classical, hellenistic, roman, protochristian, byzantine, etruscan periods, and megalithic cultures in Mediterranean region, ■ early arab cultures, ■ mythology & archaeology, ■ biblical archaeology, ■ egyptian and middle eastern archaeology ■ natural sciences applied to archaeology (archaeometry): methods and techniques of dating, analysis, provenance, archaeogeophysical surveys and remote sensing, geochemical surveys, statistics, artifact and conservation studies, ancient astronomy of both the Old and New Worlds, all applied to archaeology, history of art, and in general the hominid biological and cultural evolution. ■ biomolecular archaeology, osteoarchaeology ■ archaeology & international law ■ palaeo‐climatological/geographical/ecological impact on ancient humans ■ archaeology and the origins of writing ■ reports on early science and ancient technology ■ cultural interactions of ancient Mediterraneans with peoples further inland. The MAA Journal: ■ includes articles written for the specialist, but including explanatory introduction for the non‐specialist in mind, ■ provides occasionally reviews of relevant techniques, applications and thematics on a local or regional basis, ■ includes research notes in updating and assessing in depth some themes of the present status of available archaeometric methods and techniques, as well as techniques in field excavation and on conservation matters, ■ disseminates information on new techniques developed and ongoing excavations taken place around the Mediterranean region, ■ encourages international discussion on the coupling between archaeology and archaeometry in their broader sense, initiating forums of discussion on the establishment of widely accepted criteria of correct approach and solution of particularly current and future archaeological problems, ■ provides professional archaeologists, local authorities and site developers with information on the usefulness of scientific methods and techniques for their execution plans, the interpretation, the preservation and the presentation of antiquities, ■ encourages and promote collaboration and develop plans for the undertaking of research and application initiatives in the field of archaeology, ■ puts forward innovative and creative ideas and enhance the Mediterranean research for the more efficient exposition and study of the EuroMediterranean cultural, heritage. EDITORIAL FORWARD BY GUEST EDITORS The papers included in this special issue of Mediterranean Archaeology and Archaeometry are a selection of double-blind peer-reviewed papers of the first International Workshop on Virtual Archaeology, Museums and Cultural Tourism – VAMCT’13, held in Delphi, Greece, September 25-28, 2013. The workshop was organized and directed by Delphi native Ioannis Liritzis, Professor of Archaeometry at the University of the Aegean, Rhodes, Greece, in valuable co-organization since its conception with Professor Maurizio Forte of Duke University (USA).The wide range of themes covered by the papers, shows that this ambitious workshop explored diverse innovations of digital technologies to a broad spectrum of challenges related to cultural heritage and tourism. Participants came from across Europe, Russia and the United States of America. One major contribution focused on a number of presentations concerning digital visualization for diverse purposes, such as integrated gaming platforms, methods to test theories related to sculptural arrangements, immersive tools for cultural tourism, and methods for collating, archiving and disseminating data. Others were concerned with how technology can enhance interaction for site-specific audiences. While several of the papers at the workshop dealt directly with Delphi itself, the archaeological site of Delphi and the imposing cliffs of mount Parnassus provided an exhilarating backdrop to the ideas and innovations presented at the workshop. One of the innovative components of the workshop was its dual focus on research and cultural tourism. While other conferences worldwide include cultural tourism conceptually, they do not explicitly see those explorations as part of their missions. The conference was held at the European Cultural Center of Delphi on the west edge of the modern town of Delphi. It was an ideal location, well-appointed and efficiently organized, with exceptionally smooth integration of various technologies. The transdisciplinary (team research) nature of papers presented in Delphi led to our expanding the number of guest editors of this volume to the six of us listed below. The 2013 workshop wishes to thank the Ministry of Culture & Sports for its support. The 2nd International Symposium on Virtual Archaeology, Museums and Cultural Tourism will again be held in Delphi, September 23-26, 2015; the website is up: http://vamct.syros.aegean.gr/2015/. Arne Flaten, Maurizio Forte, Thomas E. Levy, George Pavlidis, Spyros Vosinakis, Ioannis Liritzis Mediterranean Archaeology and Archaeometry, Vol. 14, No 4, pp. 1-10 Copyright © 2014 MAA Printed in Greece. All rights reserved. AUGMENTED REALITY FOR ARCHAEOLOGICAL ENVIRONMENTS ON MOBILE DEVICES: A NOVEL OPEN FRAMEWORK Dr Ioannis Deliyiannis1, Dr Georgios Papaioannou 2 1 2 Interactive Arts Lab, Department of Audio and Visual Arts, Ionian University, Corfu, Greece Museology Lab, Department of Archives, Library Science and Museology, Ionian University, Corfu, Greece Received: 22/11/2013 Accepted: 15/07/2014 Corresponding author: Dr Georgios Papaioannou (gpapaioa@ionio.gr) ABSTRACT The wide availability of networked mobile devices provides a reliable platform for the development of the so-called communication engine for museums and cultural tourism. This research presents and discusses a novel open framework, which can be employed to augment the visitor’s experience and present targeted information in a personalised audio-visual interactive manner on users’ personal mobile devices. The proposed approach employs state of the art augmented-reality technologies enabling users to sample the information through the use of their personal mobile devices. Instead of using tagging systems such as visible quick response (QR) markers, users are directed to 1) stand on specific appropriately marked information points, 2) scan the area with their appropriately configured mobile device, and 3) access specific geographical or artefact-based ontologies that may include digitally restored buildings in 3D, audio-visual information on specific artefacts and/or other information of interest with directions to access other information points. The proposed framework may be employed at varying levels of complexity, enabling the development of archaeological edutainment scenarios and games. The use of the proposed technology has multiple advantages, such as: 1) highly-specialised hardware is not required, 2) devices can function in both open and closed spaces, 3) the quality of presentation adapts according to the device used, and 4) further information may be accessed as full interaction is supported. In this paper we review the literature and present technologies and related research that may be employed for the presentation of archaeological information. We also describe the proposed open framework, followed by a presentation of a sample application, --Additional uses are proposed in our conclusions. KEYWORDS: augmented reality, networked mobile devices, archaeological sites, museum technologies, virtual tours, interactive multimedia 2 DELIYIANNIS & PAPAIANNOU 1. INTRODUCTION This work targets the development of real-life paradigms based on the communication engine for museums and cultural tourism. In the last two decades, Information and Communication Technology (ICT) has shaped museums’ and cultural tourism communication schemes, digital catalogues, databases, websites, online exhibitions, interactivity, virtuality and an increasing wealth and diversity of devices and interfaces characterise ICT communication in cultural heritage settings. This has led the use of various multimedia technologies, which have been actively employed to enhance the user experience, inform, educate and provide information in an entertaining fashion. These technologies include various tools such as video projections, multimedia rooms / applications and wide-screen TVs (Economou, 1998) (CD / DVD players using headphones, touch screens/PC stations and interactive kiosks (E Hornecker & Stifter, 2006) audio and/or digital tour-guide systems with headphones (E Hornecker & Bartie, 2006) museum robotics (Reitelman & Trahanias, 2000; Thrun et al., 1999), and recently mobile smart-phones (Brown & Chalmers, 2003; Yannoutsou, Papadimitriou, Komis, & Avouris, 2009) tablets and Personal Digital Assistant (PDA) devices (Tesoriero, Lozano, Gallud, & R., 2007), while recent examples present multiple innovative examples and approaches (D. Vanoni, M. Seracini, & Kuester., 2012; T. E. Levy et al., 2012). Research projects, papers and special congresses (e.g. CAA 1, Museums and the Web 2, VAST 3, CHNT 4) have presented, analysed and evaluated ICT applications in cultural heritage in relation to museum settings, visitors’ experience and visitors’ parhttp://www.caa2013.org/drupal/Home (accessed 01/09/2013). 2 http://www.archimuse.com/conferences/ mw.html (accessed 01/09/2013). 3 http://www.vast2012.org/ (accessed 01/09/2013). 4 http://www.stadtarchaeologie.at/ (accessed 01/09/2013). ticipation in exhibition content and design (Adams & Moussouri, 2002; Economou, 1998; Keene, 1998; Lepouras & Vassilakis, 2004; Papaioannou & Stergiaki, 2012; Project, 2008; Pujol-Tost, 2011; Roussou, 2012; Sylaiou, Liarokapis, Kotsakis, & Patias, 2009). Our intention is to contribute to both the relevant discussion and to the standardization of interactive multimedia technologies. Today, the wide availability of multimedia-enabled handheld and networked devices such as mobile phones and PDA’s enables the development of advanced working paradigms that may be used to recognise exhibits and by linking to their virtual ontologies can provide information in an interactive and adjustable manner (De Paolis, Aloisio, Celentano, Oliva, & Vecchio, 2011; Ioannis Deliyannis, Giannakoulopoulos, & Varlamis, 2011; Ronchi, 2009; Sauvé, 2009; Torrente, Mera, Moreno-Ger, & Fernández-Manjón, 2009). The way that information is triggered and accessed is a key characteristic to our research as our previous work in complex interactive systems has shown that there may be multiple uses for the same ontology for different user groups (Ioannis Deliyannis, 2007, 2011, 2012b; Ioannis Deliyannis et al., 2011). In order to comprehensively investigate the issues in hand, we examine the process from three distinct yet complementary perspectives: user, content and technology, as shown in Figure 1. 1 Figure 1 Interrelation between users, content experts and developers © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 1-10 AUGMENTED REALITY FOR ARCHAEOLOGICAL ENVIRONMENTS Interaction with content is a key issue commonly introduced in the development of interactive multimedia systems, as it is used as a customisation tool, enabling content interrogation. As a result augmentation is a problem that must be examined from various viewpoints, as it poses multiple issues that need to be addressed simultaneously in order to develop a system that covers both the content presentation demands and user requirements. In that respect, multimedia technologies possess a secondary role, as they are adjusted appropriately in order to provide the content access platform that serves the content and satisfies the user demands. In the following sections we discuss augmentation, propose a novel framework that utilizes augmented-reality technologies and present a number of case studies developed to demonstrate the proposed approach in practice. The final section formalizes the framework and discusses customization and adaptability issues that often arise in the development of similar systems. In the conclusion, a number of suggestions are listed describing innovative uses of the proposed framework. 2. ARCHAEOLOGICAL CONTENT, INTERACTIVE AUGMENTATION AND CONTENT EXPERTS In the literature, it is common to categorise interactive systems by contrasting their dynamic capabilities offered to the user (Nardelli, 2010; Trifonova, Jaccheri, & Bergaust, 2008). Use of the term “interactive” within the archaeological context implies that the end-user of such a system may access information in a multimodal manner and the system should provide capabilities that allow users to view, investigate and explore the stored content in multiple modes. In that respect, augmentation of content is classified as a complex task, particularly when full user-content interaction is supported by the end-system (Ioannis Deliyannis, 2012a; I. Deliyannis, 2013). “Edutainment” is a characteristic research area where one may find increased 3 information on the development of interactive augmentation (De Paolis et al., 2011; Ioannis Deliyannis et al., 2011; Encarnação, 2007; Green & McNeese, 2007; Oh & Woo, 2008; Ronchi, 2009; Sauvé, 2009; Torrente et al., 2009). Archaeological content possesses a number of characteristics that are taken into consideration in the design of the proposed framework: archaeological artefacts, ruins, buildings and areas are often well represented from the information perspective. They are concisely categorised, documented and depicted, offering an information domain that apart from the addition or correction of facts in its description, requires little maintenance in time. As a result, this particular content is considered as ideal for interactive augmentation, a term used in this work to describe the ability of the user to access archaeological content interactively. Augmentation has evolved and today a plethora of platforms are available for experimentation in the development of new application systems. QR marker technologies detect and decode visual markers in order to link to information. Various open source libraries and multimedia-authoring software (Processing, 2013) may be freely programmed to create high-level interactive multimedia systems. The approach where object identification relies on marker recognition introduces various disadvantages, as the existence of the marker itself may spoil the aesthetics of the environment. The introduction of augmented reality systems based on real-life image recognition provides a much more flexible identification method. Junaio and Aurasma are two characteristic web-based platforms that provide the basic functionality required for such applications, while users may freely experiment with the platform using their own devices. As a result we have already developed a number of edutainment case studies in order to experiment with the limitations of such technologies, as part of the “Edutainment” course in the Department of Audio and Visual Arts, Corfu Greece. Students were asked to design and develop an interactive scenario © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 1-10 4 DELIYIANNIS & PAPAIANNOU using their platform of choice, a process that created various case studies each of which provided a novel interactive approach. These according to their intended purpose and functionality, these include augmented-reality based hidden object games, navigation & storytelling scenarios where video is triggered by artefacts and signs, interactive scenarios and puzzles. In order to enable content-experts to appreciate the content organisation requirements, we contrast the organisation of a physical museum collection to that of a vertical portal of information often referred to as a “Vortal”. When information access is contrasted at the top level, it is evident that both structures share solid content categorisation, enabling visitors and users to either scan through the content, or locate an area of interest and dig in to the information provided. Conversely, at the lowest representation level, physical exhibits and virtual ontologies can be mapped on a oneto-one basis, enabling direct access to information. The information access route is one of the most important characteristics of such a system that is often overlooked by developers, as the structure usually follows the thematic sectioning followed in the physical world. Virtual museums employ multiple routes and provide content adjustment capabilities. An example of such a route may enable the same exhibits to be presented using adapted information for different target-age visitors such as different members of a family. Another important feature is the capability to identify, locate and examine specific artefacts and their development through time. Examining the development of pottery designs through time can the basis of an educational visit to a museum where each student group can complete a different research scenario and present that in class. This feature can convert a visit from a static presentation of knowledge to an interactive investigation tool that may be used freely by the user enabling content-interrogation (Ioannis Deliyannis, 2011). A third feature would be to over impose digitally drawings, 3D re- constructions and other information regarding artefacts and open spaces, enabling the user to view a reconstructed perspective of the original item or space. The typical order in which the development of such a system is addressed from the software engineering perspective may be represented as a cyclic process (spiral model): first the content experts organise the collection items thematically in sections and provide an order of presentation for each section if that is necessary. Then the developers provide the platform that supports information access to the user and add the information. Finally, users test the system and provide feedback that triggers further advances in both content and technologies involved enabling further developmental cycles to be implemented. As the user plays leading role in the process, the research problem here is to provide the developmental flexibility that permits dynamic adjustment of the presentation route based on content-expert guidelines and scenarios,. This requires careful context analysis in order to allow content experts the flexibility to create adaptive multimedia presentations using the same markers for multiple purposes. In computer technology, this is analogue to dynamic URI where the same link (marker) when used by different user groups leads to varying scenarios, depending on certain conditions that may include earlier user choices, presentation scenario, visual marker availability, time-based events etc. To view this through an example, take a family visiting a museum that uses the augmented-reality applications developed for this purpose. Each family member accesses the tour by first filling in an electronic questionnaire (age-range, personal interests, specific keywords). This input enables the presentation to either adjust its content appropriately, or direct the user to specific sections of the museum. In the first case the main tour may offer full information to adults, while the children receive particularly adjusted content for their age-range. The indication of special interests may be used to focus the presentation on a specific © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 1-10 AUGMENTED REALITY FOR ARCHAEOLOGICAL ENVIRONMENTS museum section or in a cross-section setting based on a theme, say for the development of pottery and designs through various civilisations. This is a feature that may also be provided as part of edutainment scenarios, where students visiting a museum can be allocated to complete specific tasks, the findings of which will then have to present in class. It is therefore important to provide a theoretical framework that supports this kind of functionality, in an attempt to develop high-order interactive content that supports varying user scenarios, while it enhances the museum experience. 3. AUGMENTED REALITY FRAMEWORK FOR ARCHAEOLOGY (ARFA) As content customisation and system adaptability are desirable characteristics for the target application-system, the proposed ARFA framework is designed to ease the interactive nature of content. For this reason, information is organised in finite triplets (visual representation, context and corresponding audiovisual content). Visual representation is a term that refers to an artifact, a collection of items or photographs taken at a specific location that are characteristic of the area. The developers consider this a trigger item that when it is detected by the mobile device held by the user, it initiates the process. It may consist of a single or multiple images, while multiple contexts can also be linked to a visual representation. Audiovisual content can be stored within a multimedia database and be accessed on demand, furnishing varying interactive scenarios and the decisionmaking procedure may either be internal or external. In the main data-structure the user attributes are stored and linked with one or more virtual representations of each artefact, while at the system level each context is linked to its audio-visual content source. The decision process algorithm evaluates user-input and presentation attributes in order to select the appropriate context that should be displayed for each 5 user. Selection of each context information results in the projection of the appropriate information, augmented with real-life footage. This augmented projection is the resulting outcome of the process presented via the device, which composes real-life imagery with audiovisual information. At a higher-level that serves the storytelling mode, there are different types of narration that may be implemented using the data structure above: from linear to dynamic-dynamic (interactive). Multiple types may also be combined in a hybrid structure termed “cruise-control” commonly employed in scientific areas featuring complex data (Ioannis Deliyannis, 2011) that allows the user at selected points of interest to deviate from a linear structure and explore the artifacts of interest. Interaction can be implemented using various modes and techniques, depending on the environment and the level at which information can be sampled: near-item proximity for small-sized findings such as coins, tools and pottery, small rooms and closed spaces containing larger items and ultimately large exhibitions and open spaces. At the small-sized item level the user can access information by pointing at the items themselves. At small rooms and closed spaces, the introduction of appropriate floor or wall-based areas indicates that scanning from that perspective, one may also access information about the room itself and the collection organization. For large rooms and open spaces, beyond the localised information, it is also possible to provide directions to other points of interest enabling physical navigation to be implemented through the device itself. To summarize the above the visual representation of objects, spaces and surrounding environment can be used as links to content without using specific markers. In addition, it is possible to utilize the same objects and environments in order to support various interactive scenarios, by employing state-of-the-art interactive multimedia technologies (Belluci, Malizia, & Aedo, 2012; De Paolis et al., 2011; Ioannis Deliyannis et al., 2011; Khan, Xiang, © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 1-10 6 DELIYIANNIS & PAPAIANNOU Aalsalem, & Arshad, 2012; Ko et al., 2011; Paramythis, Weibelzahl, & Masthoff, 2010; Rautaray & Agrawal, 2012; Schraffenberger & Heide, 2012; Tahir, 2012; Xia, 2011). These encapsulate the visual representation of the trigger and link all related audiovisual content instances. Successful recognition of the object and selection of the content triggers a presentation instance, at a pace the user can follow. This renders the mobile device as an investigation tool. A networked multimedia database may be employed to store and access the multimedia information structure, while decisionmaking may be implemented either remotely or locally on the networked device. In any case, technologies such as the new HTML5 or ready-made platforms such as Aurasma, or Junaio may be employed to fulfill the interaction requirements. The following chapter presents a number of reallife factors that are commonly introduced in case studies. Figure 2 Two ARFA framework context triplets, and their augmented projections triggered by the initial visual representation and decision routine 3. REAL-LIFE CASE STUDIES The main purpose of this section is to define the practical application framework for archaeological use and inspire further applications. First the collection of a typical archaeological museum is examined and a number of possible uses of such technology are proposed. The collection of the archaeological museum of Igoumenitsa in Thesprotia, Greece that is organized into five exhibition units according to their website 5: archaeological - historical retrospect (items: Stone tools, hand made clay vessels, Mycenaean clay and metal objects, mobile findings of the Geometric, Archaic, Late Classic - Hellenistic, Roman and Byzantine periods), settlements of historical times (items: Architectural members and mobile findings from the ancient settlements of Elea, Gitana, Doliani and Ladochori), public life (items: Signs, coins, sealed pottery handles, clay sealings, weights and measures, mobile findings from public buildings, weapons, votives from shrines), private life (items: Tools of various professions, toys, musical instruments, loom components and pottery products, bathtubs, beautification paraphernalia, jewelry) and burial customs (items: Authentic graves and skeletons, burial offerings). From the hardware perspective, the development of an interactive presentation environment for the museum where visitors use their own devices in order to explore the collection requires the availability of networking in all areas and the availability of a local multimedia & web-server used to store and provide access to all the media information. At the information forefront, the museum holds a database containing information about all the findings that needs to be made available over the local network, presenting a unique link to the appropriately presented information for each item, covering the ARFA information triplet demands: (visual representation, context an audiovisual content). Beyond accessing the information triplet, the use of Aurasma is proposed as a platform to allow interaction with content. The information provided covers interaction with artifacts, hence there is the need http://odysseus.culture.gr/h/4/eh41.jsp?obj_ id=17461 (accessed 01/09/2013). 5 © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 1-10 AUGMENTED REALITY FOR ARCHAEOLOGICAL ENVIRONMENTS to provide further directional information. This may be realized via the creation of location-based information triplets for each room. These are implemented by capturing specific segments of the room’s panoramic view from a designated location. This specific location should also be marked using floor markings (stickers or signs) that contain also instructions and the informationaccess instructions. Single item and area recognition are supported. The next case study introduces the same technology into the open archaeological space of Paleopolis, Corfu, Greece. Information including historical photographs, 3D reconstruction and audio-based narration and WWW-links to electronic material may be employed to augment the user experience and present location-based archaeological information. In order to demonstrate this process in practice we have developed a test-application featuring a small number of locations. Figure 3 displays the representation of one point of information within Aurasma Studio the webbased component of the application supporting our case study. The highlighted (masked) area aids the recognition process, while the developers can train the system to recognise alternative perspectives and camera angles. Figure 3 Selection of the actual area of interest using a polygon. The trigger image shown in Figure 3 when recognized by the application is programmed to display a number of items, 7 one of which is shown in Figure 4. For this instance real-life imagery is linked to historic images with appropriate titles. References should be again in single column format. Use Reference Item style for references (similar to normal without indentation and with hanging 1cm). Alternatively, content experts are presented with the option to display video footage about the history of the area or composite presentations, which may utilize mixed audiovisual media. Our experiments have shown that video-based guiding such as documentary clips, digital representations and archaeological presentations in vivo are recommended for similar settings. Figure 4 A historical photograph of the ruins. 3. CONCLUSIONS This work aims to enable archaeologists employ augmented reality technologies in order to cover their particular content presentation requirements and furnish the communication engine for museums and cultural tourism with novel technological features. We introduced the ARFA framework while a number of design and development issues are presented and discussed. These include the information triplet, which unifies and links multimedia content to the trigger marker image (visual representation); various narrative options that are available for different presentation settings including include closed and open spaces; and the idea of physical information-points where visitors may utilise to receive location-based information by exploring their surrounding space. The fact that most ar- © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 1-10 8 DELIYIANNIS & PAPAIANNOU chaeological collections are well documented and the data are accessible electronically is of high importance, while in many cases existing multimedia content, 3D models and documentaries and may be edited and used directly in the target application. The framework refers to information linking and it may be combined with open access platforms or proprietary augmented reality applications like Aurasma and Junaio, which offer a cost-free developmental test bed. Our further research directions target the development of an application that unifies the information in a Geographical Information System-like environment, enabling also web-based access to the content for those who wish to virtually visit the archaeological sites and collections. REFERENCES Adams, M., & Moussouri, T. (2002). The Interactive Experience: Linking Research and Practice., Interactive Learning in Museums of Art and Design: An International Conference. Victoria & Albert Museum, London. Belluci, A., Malizia, A., & Aedo, I. (2012). Towards a framework for the rapid prototyping of physical interaction. Paper presented at the Fourth International Workshop on Physicality, BCS HCI 2012 Conference, University of Birmingham, UK. Brown, B., & Chalmers, M. (2003, 14-18 September 2003). Tourism and mobile technology. Paper presented at the Eighth European Conference on Computer Supported Cooperative Work (ECSCW 2003), Helsinki, Finland. D. Vanoni, M. Seracini, & Kuester., F. (2012). ARtifact: Tablet-Based Augmented Reality for Interactive Analysis of Cultural Artifacts, IEEE International Symposium on Multimedia (ISM). De Paolis, L. T., Aloisio, G., Celentano, M. G., Oliva, L., & Vecchio, P. (2011, 25-27 April 2011). A simulation of life in a medieval town for edutainment and touristic promotion. Paper presented at the Innovations in Information Technology (IIT), 2011 International Conference on. Deliyannis, I. (2007). Exploratory Learning using Social Software, Cognition and Exploratory Learning in the Digital Age (CELDA). Algarve, Portugal. Deliyannis, I. (2011). Interactive Multimedia Systems for Science and Rheology: Interactive Interrogation of Complex Rheological Data Using Media-Rich Adaptive Multimedia Technologies. Germany: VDM Verlag Dr. Muller. Deliyannis, I. (2012a). From Interactive to Experimental Multimedia. In I. Deliyannis (Ed.), Interactive Multimedia (pp. 3-12). Rijeka, Croatia: Intech. Deliyannis, I. (2013). Sensor Recycling and Reuse. International Journal of Sensor and Related Networks, 1(1), 8-19. Deliyannis, I. (Ed.). (2012b). Interactive Multimedia. Rijeka, Croatia: InTech. Deliyannis, I., Giannakoulopoulos, A., & Varlamis, I. (2011). Utilising an Educational Framework for the Development of Edutainment Scenarios, 5th European Conference on Games Based Learning. Athens. Economou, M. (1998). The Evaluation of Museum Multimedia Applications: Lessons from Research. Museum Management and Curatorship 17(2), 173-187. Encarnação, J. (2007). Edutainment and Serious Games – Games Move into Professional Applications. In K.-c. Hui, Z. Pan, R. Chung, C. Wang, X. Jin, S. Göbel & E. Li (Eds.), Technologies for E-Learning and Digital Entertainment (Vol. 4469, pp. 2-2): Springer Berlin / Heidelberg. Green, M., & McNeese, M. N. (2007). Using Edutainment Software to Enhance Online Learning. International Journal on E-Learning, 6(1), 5-16. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 1-10 AUGMENTED REALITY FOR ARCHAEOLOGICAL ENVIRONMENTS 9 Hornecker, E., & Bartie, P. (2006). Technology in Tourism: Handheld Guide Systems and Museum Technologies, Technical Report. Human Interface Technology Laboratory: University of Canterbury. Hornecker, E., & Stifter, M. (2006). Learning from Interactive Museum Installations About Interaction Design for Public Settings, OzCHI 2006. Sydney, Australia. Keene, S. (1998). Digital Collections: Museums and the Information Age. Oxford. Khan, W., Xiang, Y., Aalsalem, M., & Arshad, Q. (2012). Mobile Phone Sensing Systems: A Survey. Communications Surveys & Tutorials, IEEE, PP(99), 1-26. Ko, A. J., Abraham, R., Beckwith, L., Blackwell, A., Burnett, M., Erwig, M., et al. (2011). The state of the art in end-user software engineering. ACM Comput. Surv., 43(3), 144. Lepouras, G., & Vassilakis, C. (2004). Virtual museums for all: employing game technology for edutainment. Virtual reality, 8(2), 96-106. Nardelli, E. (2010). A classification framework for interactive digital artworks. Paper presented at the International ICST Conference on User Centric Media, Palma, Mallorca. Oh, S., & Woo, W. (2008). ARGarden: Augmented Edutainment System with a Learning Companion. In Z. Pan, A. Cheok, W. Müller & A. El Rhalibi (Eds.), Transactions on Edutainment I (Vol. 5080, pp. 40-50): Springer Berlin / Heidelberg. Papaioannou, G., & Stergiaki, A. (2012). Students as Co-Curators in the Virtual Museum of Folk Musical Instruments for Children: Roles, Rules and Realities. International Journal of Heritage in the Digital Era, 1(4), 631-645. Paramythis, A., Weibelzahl, S., & Masthoff, J. (2010). Layered evaluation of interactive adaptive systems: framework and formative methods. User Modeling and UserAdapted Interaction, 20(5), 383-453. Processing. (2013). http://www.processing.org. Retrieved 22 November, 2008 Project, E. (2008, 16-18 October 2008). Proceedings: International Symposium on “Information and Communication Technologies in Cultural Heritage”, The University of Ioannina. Pujol-Tost, L. (2011). Integrating ICT in Exhibitions, Museum Management and Curatorship. 26, 1(63-79). Rautaray, S., & Agrawal, A. (2012). Vision based hand gesture recognition for human computer interaction: a survey. Artificial Intelligence Review, 1-54. Reitelman, A., & Trahanias, P. (2000). Prospects of Museum Robotics in Europe, public project deliverable of the TOURBOT project: Interactive Museum Tele-presence through Robotic Avatars. Ronchi, A. M. (2009). Games and Edutainment Applications. In eCulture (pp. 375-378): Springer Berlin Heidelberg. Roussou, M. (2012). Multimedia Applications in Archaeology and Digital Interpretation. In N. A. Silberman (Ed.), The Oxford Companion to Archaeology (2 ed., pp. 1507). Oxford. Sauvé, L. (2009). Design Tools for Online Educational Games: Concept and Application. In Transactions on Edutainment II (pp. 187-202): Springer-Verlag. Schraffenberger, H., & Heide, E. (2012). Interaction Models for Audience-Artwork Interaction: Current State and Future Directions. In A. Brooks (Ed.), Arts and Technology (Vol. 101, pp. 127-135): Springer Berlin Heidelberg. Sylaiou, S., Liarokapis, F., Kotsakis, K., & Patias, P. (2009). Virtual museums, a survey and some issues for consideration. Journal of Cultural Heritage, 10(4), 520-528. T. E. Levy, C. A. Tuttle, M. L. Vincent, M. Howland, A. M. R., V. Petrovic, & Vanoni, D. (2012). The 2012 Petra Cyber-Archaeology Cultural Conservation Expedition: Temple of the Winged Lions and environs. Antiquity, 87(335). © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 1-10 10 DELIYIANNIS & PAPAIANNOU Tahir, M. (2012). Reality Technologies and Tangible Interaction: A brief guide to tangible user interfaces, augmented reality and related technologies: LAP Lambert Academic Publishing. Tesoriero, R., Lozano, M., Gallud, J. A., & R., P. V. M. (2007). Evaluating the users’ experience of a PDA-based software applied in art museums, 3rd International Conference on Web Information Systems and Technologies (WEBIST 2007). Barcelona, Spain. Thrun, S., Bennewitz, M., Burgard, W., Cremers, A. B., Dellaert, F., Fox, D., et al. (1999). MINERVA: a second-generation museum tour-guide robot. Paper presented at the IEEE International Conference on Robotics and Automation. Torrente, J., Mera, P. L., Moreno-Ger, P., & Fernández-Manjón, B. (2009). The Relationship between Game Genres, Learning Techniques and Learning Styles in Educational Computer Games. Transactions on Edutainment II, Lecture Notes in Computer Science(5660/2009), 1-18. Trifonova, A., Jaccheri, L., & Bergaust, K. (2008). Software engineering issues in interactive installation art. Int. J. Arts and Technology, 1(1), 43-65. Xia, S. (2011). Interactive Design Research of Multimedia Courseware in Teaching. Advanced Materials Research, 271, 1472-1477. Yannoutsou, N., Papadimitriou, I., Komis, V., & Avouris, N. (2009). "Playing with" museum exhibits: designing educational games mediated by mobile technology, 8th International Conference on Interaction Design and Children (pp. 230-233). © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 1-10 Mediterranean Archaeology and Archaeometry, Vol. 14, No. 4, pp. 11-16 Copyright © 2014 MAA Printed in Greece. All rights reserved. PUBLIC VIRTUAL PRESENTATION OF ARCHAEOLOGICAL MATERIALS: THE NOTES FROM RUSSIA Daria Hookk The State Hermitage Museum, 190000 Dvortsovaya embankment, 34 Dvortsovaya square, St Petersburg, Russia Received: 28/11/2013 Accepted: 10/08/2014 Corresponding author: Daria Hookk (hookk@hermitage.ru) ABSTRACT The history of museum informatics in Russian Federation can be divided in stages according to the use of various computer technologies. At the same time nobody takes in account information technologies. First of all the museum cataloguing had aim to put information in order and make it accessible. Public presentation of museum collections and results of scientific research required new technologies different from databases. The uptoday solutions are mostly based on web-technologies. Information goes to user. According to the Gartner Hype Cycle methodology, there are five key phases in every information technology’s life cycle. First, there are one to two years of announcement, followed by testing and risk evaluation. Then, there are two to three years for inflated expectations linked to the implementation and usage of the technology to peak. Finally, this is followed by a recession and a negative slope. It is curious but just the archaeological collections always was in focus of case studies for the museum informatics. They started by scientific descriptions for the special aims of archaeological research using numerical methods. Then a database for the museum catalogue was developed, and finally we continue to look for the modern techniques for the virtual reconstructions and public presentation of our collections to wide public. Sometimes the necessity to avoid expensive hardware results in the good solutions and possibility to find balance between computer and information technologies. Now, for us the problem consists in usability of the information. Our visitors are in need of data and museum specialists are not ready to apply modern techniques for their purposes. KEYWORDS: museum informatics, dissemination, Gartner Hype Cycle methodology 12 DARIA HOOKK 1. ARCHAEOLOGICAL COLLECTION AS A CASE-STUDY According to the Gartner Hype Cycle methodology, there are five key phases in every information technology’s life cycle. First, there are one to two years of announcement, followed by testing and risk evaluation. Then, there are two to three years for inflated expectations linked to the implementation and usage of the technology to peak. Finally, this is followed by a recession and a negative slope. In order to illustrate the application of information technologies in Russian museums the development of web-sites, CD-disks, infokiosks and digital catalogues was explored [Hookk 2005]. There are two aspects: the technology became out of fashion or simply old. We used an example of the State Hermitage Museum, where the first department of museum informatics in Soviet Union was organized by Jakob Sher in 1981 (Sher, 1978). Since that time the problem of transition from one technology to another and data safety has been actualised. It is curious but just the archaeological collections always was in focus of case studies for the museum informatics. They started by scientific descriptions for the special aims of archaeological research using numerical methods. Then a database for the museum catalogue was developed, and finally we continue to look for the modern techniques for the virtual reconstructions and public presentation of our collections to wide public (Hookk 2012). 2. FROM CATALOGUING TO DISSEMINATION The history of museum informatics in Russian Federation can be divided in stages according to the use of various computer technologies. At the same time nobody takes in account information technologies. Being influenced by example of Robert Chenhall (Chenhall, 1975), first of all the museum cataloguing had aim to put information in order and make it accessible. Several ages were spent without equip- ment and all descriptions were only theoretical. Some scientific problems were solved on computer PDP-11 and then certain data in 1990s were transmitted through Robotrons to the first PCs. Figure 1 A data model of museum collections data management system “Atlas” includes more than 400 parameters of description (photo by D.Hookk) The attempts to create general museum database for collection management could be considered successful at that time. The only one problem exists till now – too many data to input and that is why nobody is ready to work for future generations. Anyway a good example of use of GIS technologies combined with database was demonstrated (Mazurkevich et al., 2005; Hookk et al., 2007). Finally, the database “Atlas” developed with the help of Oracle technologies in 1998 was proposed for the archaeological collection management and it is used still. Public presentation of museum collections and results of scientific research required new technologies different from databases. There was a period in the beginning of XXI century, when museum CDdisks and sensor kiosks with Flash animations were on fashion. The up-today solutions are mostly based on webtechnologies. Information goes to user. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 11-16 PUBLIC PRESENTATION OF ARCHAEOLOGICAL MATERIALS 3. CRETERIA OF AVAILABILITY AND RELEVANCE Sometimes the necessity to avoid expensive hardware results in the good solutions and possibility to find balance between computer and information technologies. For example, a sensor kiosk is associated with onetime require (like, pay and go), it does not suit for a guided tour through the museum. Many things can be find simple in internet by Google (e.g. biography of artist). Audio-guides are useful for a single visit at the permanent exhibition. Museum has no possibility to do them for temporal exhibitions. QR-code on label requires free Wi-Fi for all visitors. Figure 2 A sensor kiosk in a War Gallery of the Winter Palace with information (photo by D.Hookk) Just now we decided to study, what kind of information is available for our virtual visitors in Internet. Virtual guide on the web-page of the museum offer possibility to get data only on 7 from 20 rooms of the archaeological exhibition. Required plug-in is not supported by modern browser. From 17 virtual courses unique topic “Scythes” on archaeology is available only in Russian. Technologies applied by developers of the application are not compatible with iPhones and iPads. In a special mobile application “Hermitage Museum” archaeological collections are presented by 5 items with invalid data taken from the old version of the web-site. Public excursions which include information on the same 20 rooms of the Winter Palace – residence of Russian tsars - and archaeological objects there are accessible for about 200 persons 13 per year, while we have about 3 million visitors. Thus we came to the conclusion that information is not full and relevant, technologies became outdate, and there is not feedback with visitors. Now, for us the problem consists in usability of the information. Our visitors are in need of data and museum specialists are not ready to apply modern techniques for their purposes. Moreover, we require an approach, easy in use and accessible for everybody, the cheapest as far as it is possible. 4. MOBILE TECHNOLOGIES Young for the young is a project organised by the Department of Archeology in order to make the different archeological or prehistoric rooms of the museum more attractive and appealing to the general public. The goal of this project is to create tours which are tailored to the individual interests of the visitors. It is important that they enhance the visit of the tourists at the museum, enabling them to view a range of artwork with a cohesive theme within a comfortable, self-guided tour. For example, if guests are interested in topics such as weaponry, clothes or ceramics, they will be able to choose a self-guided journey dedicated to these specific themes. Each selected artefact will be presented by a short historical, original and authentic text that will focus on several unique or curious facts. This non-traditional approach creates a comfortable, welcoming and adapted environment to discover the different rooms. Obtaining such emotionally-tinged information draws the visitor into the culture of the Hermitage and its complex and diverse world. The visitors will not be expected to follow a specific tour so as to view the objects; they will be free to discover each item of the topic at their own pace and will. These tours will be accessible on the museum website so as to allow visitors to download and read them on their personal technological devices. The final product - a PDF file - will thus be authorized on mobile devices such as tablets, smartphones © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 11-16 14 DARIA HOOKK and e-readers. A selection of files will be available on the website: the visitor will find the title of the topic, an abstract of its content, the estimated duration of the tour and the available languages. The possibility to share information with our visitors over the internet lies in the very ideology of personal networking technology (Charter of human rights…, 2012). Figure 3 Various forms of data implementation: plans, advisor, sensor kiosk, audio-guide rent and mobile application (photo by D.Hookk) According to their taste, the participants of the project will gather information about 10-20 objects (for example, arrows, pots, jewels, statues and so on). Each artefact will be located on the map of the archeological area and will be presented in an attractive way: a short informative text, a poem, a personal impression, etc. By participating in this project, the volunteers are given the opportunity to share their “visitors’ point of view” and their personal knowledge to the museum staff. Thanks to this new input, the museum’s content can be shown from an original and different perspective. As to the specialists and employees of the museum, they will also participate into the making of the project by providing technical information and by sharing their knowledge of the museum’s history. The design of a “topic template” will be developed by volunteers and computer designers. Furthermore, a nonexhaustive list of topics will be proposed and expanded by future participants - the more, the better. In addition, the data provided by the authors of the project will be protected under the domain of the museum’s website. As stated in its name, this project is oriented toward the younger visitors of the museum. Its audience includes not only those who come in search of information from our vast collection, or for the pleasure of viewing such masterpieces on display, but also those who might choose a profession in a related field such as art history. Therefore, the goal of this project is to create tours which are tailored to the individual interests of these visitors. These tours would be found in the form of files, accessible on the museum website. It is important that they enhance the visit of tourists at the museum, enabling them to view a range of artwork with a cohesive theme within a comfortable, self-guided tour. For example, if guests are interested in fashion, then with the help of their personal mobile device they can choose a self-guided journey to exhibits with this theme. Each exhibit would be accompanied by a short, original and emotional text which reveals several unique or curious facts. Consequently, this ensures the visitor is adapted and oriented within the museum, creating a comfortable, welcoming environment. Obtaining such emotionally-tinged information draws the visitor into the culture of the Hermitage and its complex and diverse world. Thus, this individualized, non- © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 11-16 PUBLIC PRESENTATION OF ARCHAEOLOGICAL MATERIALS traditional approach to visitor service creates a positive impression on visitors to the museum. Participants in the project will gather information about any 10-20 objects (for example, cats in paintings, cats in sculpture, cats in the palace museum and so on) according to their taste, creating a response to the object (this could take the form of a short, informative piece about the object, a poem, a personal impression…) and marking the location of each on a map of the museum. This mini-tour does not strictly regulate the order of visiting each point of interest on the map, but prompts visitors by numbering each object. It also indicates the amount of time necessary to complete the tour. Consultation during the creation of the points of interest will be provided by the specialists and employees of the museum, who will also help optimise the route that the tour takes. The depiction of the tour route on the map of the museum and the development of a template for the file will be created by volunteers and computer designers. The final product- a PDF filewill be accessible on authorized mobile devices such as tablets, smartphones and ereaders. A selection of files will be available on the website with an indication of the theme of the tour, an abstract from the text, the length of the tour as well as the language. The more options the better, with the theme “On taste and colour”. The possibility to share information with our visitors over the internet lies in the very ideology of personal networking technology (Charter of human rights… 2012). In addi- 15 tion, the data provided in the author’s publication would be protected under the domain of the museum’s website. Through participation in the project, it is possible for visitors to display their knowledge and express themselves and their individuality. From our standpoint, looking at our exhibits from a fresh point of view allows us to see them in a different light. During spring vacation, the children’s contest “Discover Hermitage!” took place. Students have proposed a route through the galleries and gather their “collection” with the help of a camera. Submissions differed by age group and theme. For children visiting from the school’s club in the centre, including numerous students studying art history in rotational classes, it would be possible to display their creations. The winning submission was chosen by its popularity- whether from a head count, or from data gathered from the contest website. The contest took into account only one “vote” for each entry ticket. Participants in the contest are advised to seek support for their submission from classmates, family and friends. In this manner, we hope to draw the attention of the adult audience too. Furthermore, this project will lead up to the celebration of the State Hermitage’s 250th anniversary in 2014, and the massive accumulation of tours will be a great present for the numerous guests of the museum. In this manner, we hope the word “Hermitage” will be accompanied not only by the word “museum”, but also by “great anticipation”. 5. ACKNOWLEDGMENTS The story telling would be impossible without consulting with my colleagues, particularly Eleonora Byzova, Andrei Mazurkevich and Maria Kuznetsova. Author is thankful to Natalia Kuznetsova, Caroline Chevillotte, Élodie Collin and Joyce Aerts from the Volunteer service of the Hermitage Museum, who helped with correct interpretation of discussion on future project. REFERENCES Chenhall R.G. (1975) Museum Cataloguing in the Computer Age. Nashville. Charter of human rights…(2012). Charter of human rights and principles on the Internet. Version 1.1. http://internetrightsandprinciples.org/site/wp- © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 11-16 16 DARIA HOOKK content/uploads/2012/12/Charter-on-Human-Rights-and-Principles-on-theInternet-Version-1-1-Draft.pdf. Hookk D.Yu. (2005) Analysis of the Information Techniques for the Museum Exhibitions. Proccedings of XIV Annual International Conference «EVA-2005 Moscow», November 28-December 2, 2005. Moscow. http://conf.cpic.ru/eva2005/eng/reports/report_487.html. Hookk D.Yu. (2012) Sources of virtual knowledge on the museum collections. Proceedings of XIV Annual International Conference «EVA-2012 Moscow», November 26-28, 2012. Moscow. https://eva.rsl.ru/en/2012/report/list/1074. Hookk D.Yu., Morozov S.V., Mazurkevich A.N. (2007) Database “MonArch” for the Keeping and Processing of the Data on Cultural Heritage. 21st Cipa Symposium, Athens. http://cipa.icomos.org/index.php?id=393. Mazurkevich A.N., Hookk D.Yu., Dolukhanov P.M., Morozov S.V. (2005) The Synthesis of GIS and Database technologies for Modeling Prehistorical Processes . 6th Archaeological Prospection. Rome. Sher J.A. (1978) The use of computers in museums: present situation and problems. Museums and Computers. Museum. №XXX, 3/4, 132–138. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 11-16 Mediterranean Archaeology and Archaeometry, Vol.14, No 4, pp. 17-24 Copyright © 2014 MAA Printed in Greece. All rights reserved. CREATING NEW LINKS AMONG PLACES THROUGH VIRTUAL CULTURAL HERITAGE APPLICATIONS AND THEIR MULTIPLE RE-USE Antonella Guidazzoli1, Maria Chiara Liguori1, Mauro Felicori2 and Sofia Pescarin3 1Cineca Interuniversity Consortium, Italy City Council, Italy 3 CNR-ITABC, Italy 2Bologna Received: 29/11/2013 Accepted: 08/05/2014 Corresponding author: Antonella Guidazzoli, Maria Chiara Liguori (visitlab@cineca.it) ABSTRACT The valorisation of the Italian cultural hotspots to enhance the tourism industry has always been at the heart of the local policies. Advanced computer applications can crucially contribute to this aim. In such a conceptual framework, an apparently "endless" series of ICT applications, sharing 3D assets, sprang into life in Bologna thanks to an Open Source background. The sequence of projects, that are going to be presented in the article, was initiated by the creation of the stereoscopic 3D medium-length movie "Apa the Etruscan". As an immediate consequence, four other projects have started, re-using some of the 3D models from "Apa ", and more are following. Our experience proves that a non possessive stance towards ones own products, speeds up and optimises time and costs, improving different end-products. The continuous transfer of models, adapted to new requirements, speeds up the productions, allowing to focus more on aesthetics, without damaging neither the source project nor the recipient, and reaching the widest audience possible. The multiplication of references can be seen even in the short term as much more fruitful than creating expensive projects closed in themselves. The rapidity of development and the increasing quality of the final product allows, hence, a distribution able to reach an increasingly wider audience, bringing the promotion of cultural heritage to a newer and higher awareness. KEYWORDS: Open Source, 3D modelling, Cultural Heritage, asset re-use 18 ANTONELLA GUIDAZZOLI et al 1. INTRODUCTION The valorisation of the Italian cultural hotspots to enhance the tourism industry has always been at the heart of the local policies. As stated by Francesco Antinucci (Antinucci, 2007), advanced computer applications can crucially contribute to this aim. In such a conceptual framework, an apparently "endless" series of ICT applications sprang into life in Bologna thanks to an Open Source background. The sequence of projects, that are going to be presented in the article, was initiated by the creation of the stereoscopic 3D medium-length movie "Apa the Etruscan", on behalf of the recently opened Museum of the History of Bologna. The movie, integrating previous researches, is now giving birth to further realisations, always aimed at attracting and engaging new audience. The 3D models, created from scratch or as an evolution of previous applications, have been released as Open Data on the site of the Municipality of Bologna (http://dati.comune.bologna.it/3d). This choice can be considered as a sort of letter of intent about the desire to promote the reuse of 3D contents for new cultural creations. As an immediate consequence, four other projects have started, re-using some of the 3D models from "Apa the Etruscan": • an archaeological narration in Machinima; • an augmented reality application coupled to an on-line video game set in the Roman era and in the Middle Ages; • an additional scene to be connected to the first part of "Apa the Etruscan" and to be show at the National Etruscan Museum of Villa Giulia in Rome, in order to explain something about the Roman Etruria; • a project supporting the UNESCO candidacy of the network of porticoes of Bologna. In all four cases, the goal is always to attract the attention of the public and visitors. The fourth project is particularly significant in this effort to virtuously recycling resources: the resources for the initiatives related to the Porticoes project, aimed at attracting new tourists, come from the tourist tax. The third project, combining the narration about northern and southern Etruria, opens a window between museums and territories. As Pascal Brackman says (Brackman, 2011), creating and multiplying links between places and events, allows, as on the web, to multiply the possibilities that people have to get in touch with these realities. Any set of references increases the visibility of the connected points; in our case the elements to be highlighted are the city of Bologna and its cultural resources. The new links are also part of an effort to show what is already known, or little known, from new perspectives. The augmented reality project, for example, will allow citizens to discover parts of Bologna unknown to most people, such as the Roman remains buried beneath the road surface. At the same time, users themselves should be able to participate in a shared narrative. In the project about the Porticoes of Bologna as UNESCO Heritage, for example, the digital platform will collect and provide visitors with a virtual environment, geo-referencing whatever documentation is available about the porticoes; users would engage with the proposed contents and add their personalised ones (Guidazzoli, 2013; Apollonio et al., 2013). Storytelling, gaming and wide participation (Anderson and Rainie, 2012), anything can converge towards the aim of raising the general level of attention to the many and beautiful cultural resources at our disposal. 2. "APA THE ETRUSCAN", AN OPEN EXPERIENCE ICT solutions applied to Cultural Heritage can really improve the valorisation of local cultural hotspots (Bellotti et al., 2013). By adopting such an approach, the creation of links and the re-elaboration of digital resources become simpler and quicker. Engaging contents, designed for citizens and tourists, can be easily deployed. Projects © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 17-24 CREATING NEW LINKS AMONG PLACES must not remain as isolated monads, but should become nodes in expanding networks. The experience started with the project "Apa the Etruscan" fits perfectly in this type of process (Fig. 1). 19 with a production pipeline based on open source resources for more than 95%. This choice has stimulated the development of a fruitful approach which considers the reuse, and the intensive exploitation of produced assets, the best way to maximize results (Fig. 2). Figure 2 Asset re-use in the “Apa” pipeline. Figure 1 “Apa the Etruscan”: the main character inside the Bolognese Archaeological Museum. "Apa the Etruscan", a 15 minute stereoscopic 3D movie about the history of Bologna, was developed as an integrated part of the tour in the Museum of the History of Bologna (Boetto Cohen et al., 2011). While waiting for research results about the impact of the new generation of museums, including the Museum of the History of Bologna, it is currently possible to get a superficial yet effective overview of the opinions expressed by visitors on Tripadvisor.it. By October, 2013 - taking into account that the museum opened in late January 2012 - on 125 reviews available, 36 make explicit reference to "Apa", and all in positive terms. The movie, consisting of 100 shots, about 20 locations, 380 blender files, 8500 textures and 1000 data files, was created by Cineca The philosophy of “re-use” was adopted while the movie was still under development: from the beginning, the creation of a repository supported the collection of the already available assets and the new ones created from scratch. "Apa" itself had started from the re-use of elements coming from previously developed projects: a situation that headed to the definition of a specific production pipeline. The principle that drove us towards these choices and set aside the fear of the hypothetical harmful déjà vu effect is the certainty that single elements, when used in different contexts, can express other meanings and other cultural contents, eliminating the danger of repetition. Once the project was concluded, it seemed a natural consequence to make the 3D models available through the OpenData website of the Municipality of Bologna under a Creative Commons 3.0 ccby-nc-sa license, asking users for attribution, sharing alike and non commercial use of the models (creativecommons.org/licenses/). In October 2013, among all data made available through the portal, one of the models produced for "Apa", related to the segment of porticoes reaching the basilica of St. Luke, was ranked 12th among the most popular con- © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 17-24 20 ANTONELLA GUIDAZZOLI et al tents and other four 3D models, again from "Apa", were located from the 30th to the 33rd place. The decision to release 3D models under Creative Commons license was also supported by a strong commitment to the Open Source vision. In particular, as we will see in the next pages, the adhesion to the community built around the Blender modelling software proved particularly fruitful, both for the excellent characteristics of the software and for the resources made available, even in this case, under various Creative Commons licenses. 3. PROMOTE LOCAL REALITIES THROUGH COMPUTER APPLICATIONS FOR CULTURAL HERITAGE: SOME EXAMPLES 3.1 Marcus Caelius, the value of memory Even before the creation of the film "Apa" was over, some of its assets have been re-used in another project: “Marcus Caelius, the value of memory”, realised by NoReal, upon suggestions by Cineca, on behalf of the Civic Archaeological Museum of Bologna (Bentini et al., 2012). “Marcus Caelius” is an experiment to assess the feasibility of creating a short educational film with Machinima technique and in an OpenSim environment (opensimulator.org). The goal is to be able to contain the costs of 3D productions by reducing at least the effort in the animation. Set in Roman Bologna in the Augustan era, it aims to highlight some findings pertaining to the collection of the Archaeological Museum by narrating the death of the Bolognese centurion Marcus Caelius in the Battle of the Teutoburg and his brother's decision to dedicate him a tombstone, currently preserved at the Museum of Bonn, Germany. The environments, set up in Open Sim, and the use of avatars, allow to "shoot" the scenes in real-time, directly in the multiverse. The story-telling gains a position of dominance with respect to the final aesthetic result. For this project, which will be shown at the Archaeological Museum of Bologna, some assets made for "Apa", such as the procedural Roman Bologna, have been exploited. At the same time, assets made for this short film, such as various objects of daily use, have become part of the production pipeline of the video game "ApaGame", presented in this article. Therefore, even before its conclusion, "Apa" had been already deeply involved in the philosophy of re-use in the framework of virtual cultural heritage applications: "Apa" had been taken advantage of researches and 3D models created by other projects and was now offering digital resources to another project commissioned by another museum. "Marcus Caelius" would have provided, in turn, 3D models to a new video game production aiming at the promotion of the city as a whole. 3.2 "Ati" and Southern Etruria "Apa" has proved extremely fertile and, after less than a year, the production of its spin-off, "Ati", started (Delli Ponti et al., 2013). "Ati" aims to enhance the collections of the National Etruscan Museum of Villa Giulia in Rome, a city other than Bologna, but telling also of Bologna during Etruscan times. The movie has been conceived as a bridge between Bologna and Rome, between Northern and Southern Etruria, borrowing directly a few minutes from the "Apa" movie. At the end of the sequence narrating the Etruscan Bologna, the version for the Museum of Villa Giulia goes on driven by another character, "Ati", briefly narrating about Southern Etruria and the temple of Veii and taking a cue from some relics preserved in the Museum itself. In this case, there is not a re-use of assets but, rather, the re-use of a rendered shot used as an introduction to something completely new. Even the narrator changes and, if on the one hand, this choice was inevitable after the death of the Bolognese singer Lucio Dalla, who had lent his voice to "Apa", on the other hand the creation of a customised mascot for the Roman museum is, obviously, a precious opportunity. This kind of solutions for the dissemination and promotion of cultural hot spots © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 17-24 CREATING NEW LINKS AMONG PLACES can also be successfully joined to merchandising activities. Moreover it is possible to foresee, in addition to the sale of a DVD, containing the 3D movie and its making-of, different gadgets such as "Ati"'s action figures in order to create not only economic outcomes, but also communication ones (Fig. 3). 21 mented reality Mobile apps, assets born for being rendered (Fig. 4). Figure 4 “ApaGame”: the Medieval market level (under construction). Figure 3 "Ati": 3D printing test for hypothetical merchandise. As for the Museum of Villa Giulia, despite its rich and wonderful collection, the number of visitors is still inadequate. Transforming the image of the museum in something more familiar and closer to common people, thanks to mascots and souvenirs, can pay dividends. Moreover, as we all know, the enormous wealth of Italian Cultural Heritage do not seem to have been adequately exploited yet (ISTAT, 2012) and this area of opportunity is ideal for flexible, cross-media ICT applications and, why not, can also give economic returns. As for "Ati", the main character is a total novelty compared to "Apa" and largely relies upon the increasingly vast resources made available by the Blender community, such as those coming from Blender Cookie (http://cgcookie.com/blender/) or Blend Swap (http://www.blendswap.com/) and from the wide selection of avatars that, inside an Open perspective, can be customized according to the various needs (Fig. 5). 3.3 “ApaGame” In the creation of the educational video game “ApaGame”, developed by CINECA, CNR-ITABC and Fraunhofer Institute, there is a more intensive use of the assets deriving from “Apa the Etruscan”. Set in the Medieval era and in the Roman Bologna during the Augustan age, “Apagame” is a test-bed within the V-Must project (www.v-must.net) to assess the practicability of passing on to different platforms, such as on-line games and aug- Figure 5 “Ati”: from the CookieFlexRig character to “Ati” The Medieval and Roman architectures used in the game are largely drawn from "Apa" (Fig. 6), while Roman objects relies upon some findings of the Archaeological Museum of Bologna reconstructed, as men- © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 17-24 22 ANTONELLA GUIDAZZOLI et al tioned before, for the "Marcus Caelius" project. "ApaGame" is designed as a small magnet for locals and tourists, transmitting, on the one hand, knowledge about the city and its daily life in the past and, on the other hand, trying to attract people to the nowadays city. Figure 6 "ApaGame": the Roman bridge level seen in Blender while setting the bounding boxes. In particular, the Augmented Reality app will allow the viewers strolling in the center of Bologna to enjoy the view of historical overviews through their own mobile devices. Roman vestiges in Bologna are scarcely considered also by the citizens themselves, who are undoubtedly more familiar with the medieval city and with an urban landscape still strongly characterized by buildings pertaining to that era. The information about the city and its daily life in the two periods will be transmitted through various cross-references between the two platforms – on-line game and mobile app. The UNESCO Porticoes Project will offer an important contribution to the Mobile app, since it will provide free Wi-Fi areas along the extension of the porticoes in the city center. ry of Bologna. Moreover, porticoes are a place for socialisation, trades, crafts, civil participation; they are a space for teenagers’ social life, a meeting point for housewives, informal clubs for intellectuals. Porticoes are the generous nest for those relationships for which Bologna is famous in the world. To give an account of this vitality is the declared challenge of the project, launched by the Municipality of Bologna in support of the nomination proposal for the UNESCO’s world heritage list. The Municipality will try to get the inclusion by asking the city to take on the project as a collective work. Not only professional work, then, but many volunteers (students of engineering and technical institutes, ICT enthusiasts, traders and craftsmen) will be called upon to contribute to the effort, while the independent associations will be offered virtual porticoes to stage their free expressiveness. The project includes, in addition to an effective management and maintenance of the porticoes, the implementation of a Web based geographic platform. The platform will collect, display and deliver 3D models within historical and artistic data related to Bologna porticoes. This feature is useful to describe and promote the whole porticoes system. The platform will be accessible through social platforms in order to allow users a direct participation by means of personal contents (photos, drawings, posts, etc..) and making available to a general public the updates from the Events section (Guidazzoli et al., 2013; Apollonio et al., 2013). 3.4 The Porticoes Project for a UNESCO candidacy The porticoes of Bologna are an excellent test for the potentialities of digital technologies. First of all, of course, they are architectural elements, reflecting an evolution lasting for a thousand years and they contain works of art, which are part of the city's heritage, and hundreds of plaques (ready to become hotspots!) telling the sto© University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 17-24 CREATING NEW LINKS AMONG PLACES Figure 7 X3DOM testing for the on-line 3rd Cloister of the Bolognese Monumental Cemetery linked to its database. The 3D models will partly come from the work done by students of the department of Architecture, University of Bologna, partly from previous projects, such as the Medieval Bologna from "Apa" and "ApaGame" or the Virtual Museum of the Certosa (http://www. certosadibologna.it). The Virtual Museum of the Certosa, undertaken by Cineca always on behalf of the Municipality of Bologna, had already contributed to "Apa" with terrain models (DTM). Other 3D models will converge towards the porticoes platform from the work of high school students and volunteers that can contribute in compliance with a set of requirements. The 3D models will be navigable online in X3D and will be linked to fact sheets (Fig. 7). 4. CONCLUSIONS The involvement of the general audience in a collective effort for the enhancement and preservation of cultural assets is an increasingly popular goal, as evidenced also by the CreativeCH project (http://www.creative-heritage.eu/). Our experience proves that a non possessive stance towards ones own products, speeds up and optimises time and costs, improving different end-products. In our case, the involvement of two institutions such as CINECA and CNR- ITABC, two museums (the Archaeological Museum and the Museum of the History of Bologna) and a local authority, in an on going collaboration that has been lasted for years, has allowed positive virtuous synergies that will now embrace even the National Etruscan Museum of Villa Giulia. Undergoing a first amendment, a previous work made by CNRITABC on Roman Bologna had become part of "Apa"; models that have been further re-used in "Marcus Caelius" and "ApaGame". Given the low budget availa- 23 ble for the development of "Apa", for example, it would never have been possible to create a movie capable of getting many awards without starting from an important core of available assets. Similarly, if "Ati" had not had access to the resources of the Blender community, the time needed to create the main character would have been at least twice and required higher economic resources, jeopardizing the entire project. The continuous transfer of models, adapted to new requirements, speeds up the productions, allowing to focus more on aesthetics, without damaging neither the source project nor the recipient, and reaching the widest audience possible. The desire to draw attention to their own contents should not close cultural institutions and local authorities in a jealous isolation. The multiplication of references can be seen even in the short term as much more effective than creating expensive projects closed in themselves. The rapidity of development and the increasing quality of the final product allows, hence, a distribution able to reach an increasingly wider audience, bringing the promotion of cultural heritage to a newer and higher awareness. 5. ACKNOWLEDGEMENT The authors wish to thank Davide Borra, NoReal; Paola Giovetti, Archaeological Museum of Bologna; Antonio Baglivo, Giosué Boetto Cohen, Chiara Bonanni, Francesca Delli Ponti, Luigi Calori, Daniele De Luca, Silvano Imboden, Fabio Negro, Rossella Pansini, Maurizio Quarta, Manuela Ritondale, Francesco Veronesi, CINECA; Luigi Virgolin, Comune di Bologna; Emanuel Demetrescu, Daniele Ferdani, Luca de Felice, Yaara Ilan, Guido Lucci Baldassari, Augusto Palombini, CNR-ITABC. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 17-24 24 ANTONELLA GUIDAZZOLI et al REFERENCES Anderson, J., Rainie, L. (2012), The Future of Gamification, Pew Research Center, May. Antinucci, F. (2007) Musei Virtuali, Laterza, Roma-Bari. Apollonio, F.I., Gaiani, M., Fallavolita, F., Ballabeni, M., Zun, Z., Guidazzoli, A., Baglivo, A., Liguori, M.C., Felicori, M. and Virgolin, L. (2013) Bologna porticoes project. A 3D repository for WHL UNESCO nomination, in Alonzo C. Addison, Livio De Luca, Gabriele Guidi, Sofia Pescarin (Eds.), Proceeding of: 2013 Digital Heritage International Congress, 28 Oct – 1 Nov 2013 Marseille, France, IEEE, ISBN: 978-14799-3169-9, Vol. I, pp. 563 – 570. Bellotti, F., Berta, R. and De Gloria, A. (2013) Virtual Heritage Le Tecnologie dell’Informazione (IT) applicate ai Beni Culturali, in «Storicamente», 9, DOI 10.1473/stor465,http://www.storicamente.org/02_tecnostoria/virtual_heritage. htm Bentini, L., Borra, D., De Luca, D., Guidi, F., Donati, C., Giovetti, P., Guidazzoli, A., Liguori, M.C., Marchesi, M., Pirotti, A. and Spigarolo, M. (2012) Marcus Caelius Project: a transmedial approach to support cultural communication and educational experiences, in proceedings of: D. Borra, P. Porietti (a cura di), MIMOS 2012, Roma. Boetto Cohen, G., Calori, L., Delli Ponti, F., Diamanti, T., Guidazzoli, A., Imboden, S., Liguori, M.C., Mauri, A., Negri, A. and Pescarin, S. (2011) Apa the Etruscan and 2700 years of Bolognese History, in ACM SIGGRAPH ASIA 2011, Posters and Sketches Proceedings, Hong Kong. Brackman, P. (2011) NOKIA-UNESCO Roundtable on Heritage, Tourism, and Sustainability, UNESCO, Paris, 14 to 15 March 2011. Delli Ponti, F., De Luca, D., Guidazzoli, A., Imboden, S. and Liguori, M.C. (2013) 3D Computer Graphics short movies for communicating cultural heritage. An open source pipeline, in Alonzo C. Addison, Livio De Luca, Gabriele Guidi, Sofia Pescarin (Eds.), Proceeding of: 2013 Digital Heritage International Congress, 28 Oct – 1 Nov 2013 Marseille, France, IEEE, ISBN: 978-1-4799-3169-9, Vol. II, pp. 325-328 Guidazzoli, A., Liguori, M.C. and Felicori, M. (2013) Open Creative Framework for a Smart Cultural City: Bologna Porticoes and the Involvement of Citizens for a UNESCO Candidacy, in Information Technologies for Performing Arts, Media Access, and Entertainment Lecture Notes in Computer Science Volume 7990, 2013, pp 58-65. ISTAT (2012) Annuario statistico Italiano, Attività culturali e sociali varie. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 17-24 Mediterranean Archaeology and Archaeometry, Vol. 14, No 4, pp. 25-33 Copyright © 2014 MAA Printed in Greece. All rights reserved. IMMERSIVE TECHNOLOGIES TO EXPLORE THE CYRENE TREASURY AT DELPHI Arne R. Flaten1, Susan J. Bergeron¹, Marcello Garofalo², C. Brandon Rudolph, and Jeffrey Case¹ 1Coastal Carolina University, Dept of Visual Arts, Conway, USA ²Gallagher & Associates, LLC, Washington, D.C, USA Received: 19/08/2013 Accepted: 19/09/2014 Corresponding author: Susan Bergeron (sbergero@coastal.edu) ABSTRACT The Ashes2Art project, an interdisciplinary undergraduate seminar offered every spring at Coastal Carolina University, has developed digital 3D models and other webbased resources related to the ancient site of Delphi, Greece since 2007. In 2012, the focus was on an archaeometric reconstruction of the 4th century Cyrene Treasury in the Sanctuary of Apollo. The following year the program expanded to explore the development of a comprehensive platform for synthesizing and integrating these disparate sources through an enhanced immersive landscape that leverages geospatial information, stateof-the-art consumer graphics, and gesture-based and audio navigation and interaction. A prototype for this natural user interface driven interactive platform was recently developed by students and faculty in the Ashes2Art program, utilizing a student-constructed 3D digital model of the Cyrene Treasury at Delphi as a test case. Using a virtual reconstruction of the Delphi landscape as a jumping-off point, users can explore the 4th century BCE site and its reconstructed monuments and cultural features. As a user approaches individual models, options allow them to explore the interiors and access interactive features to delve further into related media, scholarly primary and secondary sources, and additional high-detail models. The platform's natural user interfaces also allows for multiple users to interact with the platform simultaneously, allowing for instructor-student interaction and collaboration. KEYWORDS: virtual models, reconstructions, Delphi, Cyrene, treasury, gesture-based, Ashes2Art 26 ARNE FLATEN et al 1. INTRODUCTION As virtual archaeology and virtual 3D reconstructions of archaeological sites and ancient structures become more commonplace, interest has turned to developing interactive platforms that can allow the exploration of these reconstructions within a larger landscape context. Such immersive virtual platforms can be designed to extend the user’s experience of the virtual landscape by embedding and linking digital scholarly information and multimedia that can illustrate the sources used to develop 3D reconstructions, animations of the reconstruction process, links to resources and data sets, photographs, and other related materials (Bergeron 2011; Harris et al. 2011). In 2012 undergraduate students in the Ashes2Art program at Coastal Carolina University completed a block-by-block digital model of the 4th century BCE Cyrene Treasury at Delphi, Greece based extensively on the archaeological reports published by the French School in 1952. That model was then exported to an intuitive, user-friendly interface for gesture-based and voice-based interaction within an immersive virtual Delphi landscape. Options allowed users to explore the interiors and access interactive features to delve further into related media, scholarly primary and secondary sources, and additional highdetail models. 2. ASHES2ART The Ashes2Art program at Coastal Carolina University in Conway, South Carolina began humbly in 2005 as an experimental, upper level undergraduate seminar offered each spring that aimed to fuse teaching with research and technology across various disciplines. From 2007 to 2009 Coastal Carolina University collaborated with Dr. Alyson Gill and students at Arkansas State University on the first stages of examining the ancient site of Delphi, Greece (Flaten 2009; Gill 2009). Faculty at both universities led students to Delphi and various PanHellenic sites (Nemea, Corinth, Olympia, Delos, Aegina) over consecutive summers to document the sites with photographs, digital panoramas, and GPS. Our collaboration was supported by the National Endowment for the Humanities in the United States, the Hellenic Ministry of Culture, the American School for Classical Studies in Athens. In addition, Dr. Elena Partida at the Delphi Archaeological Museum provided invaluable assistance to the project. We have reported on various aspects of the Ashes2Art project regularly in journals and conference proceedings (Flaten 2009; Gill 2009; Gill and Flaten 2008), and our students have presented about our theoretical framework, our successes, and our failures at conferences in Australia, Beijing and Washington (Garofalo 2013; Rudolph 2013). Fundamental to Ashes2 Art is the concept that all materials are designed, built, coded, and implemented by undergraduate students, including digital models, web design, photography, digital panoramas, lesson plans, research essays, animation, and educational videos. Three principles have guided our study: 1) Precision does not imply accuracy; 2) Uncertainty is a crucial component of knowledge; 3) Questions are more important than definite answers. As the project has grown, so has the number of buildings we have digitally reconstructed, the level of complexity in each model, and the methods employed to explore, archive, disseminate, and interact with information. In addition to the reconstructed plunge bath and gymnasium by undergraduate students at Arkansas State in 2007 (Gill and Flaten 2008), Coastal Carolina University undergraduate students modeled the Tholos of Athena Pronaia, the Athenian Treasury, and the Temple of Apollo, among others between 2007 and 2011 (Flaten 2009). These last three structures are perhaps obvious choices for virtual reconstruction: well-known to tourists and scholars, and well documented in most cases. The Siphnian Treasury might have been a logical choice for the next project, as the frieze and caryatids are standard to © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 25-33 IMMERSIVE TECHNOLOGIES AT DELPHI most surveys of Greek architecture and sculpture. We chose, instead, to investigate something more esoteric and turned in 2012 to the Cyrene Treasury. 2.2. Treasury of the Cyrenes The digital reconstruction of the Treasury of the Cyrenes at Delphi was fascinating for several reasons: 1) it is largely absent in the scholarly literature; 2) its half columns are atypical of treasury design; and 3) the mathematical principals and implications behind the structure’s design are disputed, as is its precise location. The Cyrene Treasury received comprehensive study in the Fouilles de Delphes, and we had access to those volumes (Bousquet and Fomine 1952). In addition, the Treasury of the Cyrenes had not, to our knowledge, been the focus of a 3D virtual reconstruction. The Treasury of the Cyrenes at Delphi was built in the 4th century BCE by the prosperous Greek colony of Cyrene, which had been founded in 630 BCE when Grinnos, the king of the island of Thera (modern Santorini), was directed by the Delphic oracle to establish a port city in Libya. The site quickly established trade with various Greek cities, and became the chief commercial center for ancient Libya (near modern Benghazi). Herodotus described the city’s founding and its history, and it appears in the Old Testament and sundry ancient texts. Cyrene was the locus for a famous school of philosophy founded by Aristippus, a student of Socrates. After the death of Alexander the Great, Cyrene became subject to the Ptolemaic dynasty in Egypt and it later was consumed by the Roman empire. The archaeological site of the city of Cyrene in Libya reveals that it included a temple of Apollo, perhaps built as early as the 7th century BCE, a temple of Demeter, a temple of Zeus Ammon, and a large necropolis. The Cyrene Treasury at Delphi was located to the East within the walls of the main sanctuary of Apollo. There is some disagreement as to which foundation 27 should be associated with this building (Bousquet and Fomine 1952; Dinsmoor 1957; Laroche 1988; Partida 2000): Dinsmoor favored foundation 302 (Atlas XIV) early in the century, but he revised his decision in his review of Bousquet’s text and concured that the Cyrene Treasury was adjacent at foundation 203 (Atlas XIII). Laroche (1988) argued that Dinsmoor’s original placement was correct and Partida (2000) agreed in her survey of Delphic treasuries. The two sites are next to each other, but faced different directions (203 faced Southeast; 302 faced Southwest) (Fig.1). Figure 1. Cyrene Treasury location, Delphi (Photo credit: A. Flaten) If we follow the argument of Laroche (as well as Partida and early Dinsmoor), the treasury was built into the northeast wall of the Delphi sanctuary and its entrance faced the Athenian Stoa and the Sacred Way (Laroche 1988; Partida 2000; Dinsmoor 1957). Its priveledged proximity to the Temple of Apollo and the Sacred Way was impressive. Laroche (1988) also argues convincingly that the structure was built between 334 and 323, revising earlier scholars’ suggestions that construction on the treasury began ca. 373 BCE, was interrupted in 338 for the Battle of Chaeroneia, and was completed shortly thereafter. Their treasury at Delphi is certainly indicative of Cyrene’s wealth, but it must be understood more specifically as an offer of thanksgiving to the Delphic oracle who di- © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 25-33 28 ARNE FLATEN et al rected that city’s establishment on the shores of North Africa in the 7th century. The Treasury of the Cyrenes at Delphi was similar in many respects to other treasuries there. It was a windowless rectangle constructed of Pentelic and Parian marble measuring 3.3m x 4.6m. The plan was relatively simple: a modified distylein-antis with engaged Doric half columns flanking the two Doric columns at the entrance. Half columns are not typical of treasuries from the period and region, but their appearance on the treasury at Delphi may represent the Cyrenes themselves. The University of Manchester Expedition to Cyrene in 1952 observed that the proliferation of half columns on numerous monuments at Cyrene, to be understood as a local symbol or emblem, was first established with the treasury at Delphi (Rowe 1956). According to Bousquet, the treasury was designed to validate or explore the mathematical theories of Theodorus of Cyrene, a pupil of Plato. Bousquet asserted that the treasury’s design demonstrated answers to several mathematical problems, including the famous “squaring the circle” conundrum. Dinsmoor’s review of the publication agreed with the attribution of the building to the Cyrenes, but was highly critical of Bousquet’s conclusions concerning its mathematical properties, as was Cook’s evaluation (Dinsmoor 1957; Cook 1954). Haspels and Plommer, on the other hand, agreed with Bousquet’s mathematical assertions (Haspels 1953; Plommer 1954). Little has been written on the topic in over half a century. Professors at Coastal Carolina University in Mathematics and Art History are planning a complete review of the various aspects of the initial argument and its rebuttal, but that is not our present focus. 2.3 Virtual Cyrene Treasury A senior undergraduate student participating in the Ashes2Art course focused on modeling the Cyrene Treasury, utilizing Trimble SketchUp. Our primary source for the model was the Le Tresor des Cyrenes (1952), authored by Jean Bousquet and illustrated by Youry Fomine, and part of the extensive Fouilles de Delphes excavation report series. Even if the foundation location of the treasury is contested, the reconstruction of the treasury by Bousquet and Fomine is by far the most comprehensive and is universally admired. We decided to individually construct each block represented in the Fouilles de Delphes report, not just the superficial dimensions and larger polygons associated with many reconstruction projects. The most important drawings for rebuilding the Cyrene Treasury were located on Plate XXXIX- Observations of the Technique (Bousquet and Fomine 1952). Those drawings provided much of the necessary information to accurately model nearly every block of the treasury, with detailed illustrations of the orthostates, metopes, triglyphs, and cornice blocks. The surprising Ionic features of the treasury--the cyma reversa and ovolo on the cornice blocks, the cavetto on the metopes separating each triglyph, and the backs of the triglyphs--were absolutely crucial to the reconstruction model. Following those drawings and the text, the initial modeling of the Cyrene Treasury was completed using SketchUp, and the model was optimized wherever possible to allow for export and use in other applications (Garofalo 2013) (Figures 2 and 3). Ashes2Art frequently begins models in SketchUp and then exports them to 3D Studio Max and MentalRay for textures, lighting, animation, etc. We plan to export the model, add textures and lighting for a flythrough video during the Spring 2014 session. The aim of virtually reconstructing the Cyrene Treasury was to focus on historical accuracy and visual authenticity, as had been the mission for the Temple of Apollo and our earlier models. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 25-33 IMMERSIVE TECHNOLOGIES AT DELPHI Figure 2. View of Cyrene Treasury model in Trimble SketchUp Figure 3. Exploded view of Cyrene Treasury 3D model. Fomine’s reconstruction drawings-sections, elevations, moulding details, and floor plans--are exquisitely detailed and address almost all the extant architectural elements based upon excavation artifacts. In total, some 39 plates of drawings were published. In some instances, however, the stated measurements of the drawings needed to be verified with scale rule measurements of those drawings...and the two sets of measurements did not always agree (some of the drawings were corrected in Bousquet’s text). This process of redundant measuring, however, ensured correct spacing of elements such as the triglyphs, metopes, mutules, regulae, guttae and cornice blocks. It was also useful in obtaining accurate angles for the column capitals, pilaster molding, cornice blocks, sima blocks, geisons, gutters and roof tiles. Based on Fomine’s drawings, the corresponding text, and our 29 measurements we concluded that the raking sima, roof tiles, gutters, spouts, and doorway moulding lacked sufficient evidence for exact modelling. Either a drawing from a crucial vantage point did not exist, or the details were unclear. Comparanda of Doric temples and treasuries of the 4th century provided useful clues to model those elements. The blocks of the raking sima presented the greatest modelling challenge due to the odd angled cuts which lock them together. Fomine’s illustrations did not provide sufficient guidance to build the complicated geometry of these elements. Looking to contemporaneous sources, the Temple of Apollo at Bassae, among others, offered a helpful parallel. Had we relied exclusively on Fomine, there would be no way to understand the engineering or to appreciate the sophistication of the design. These types of problems do not manifest themselves unless one builds each stone or component individually, and they necessitate accurate (or at the very least, workable) solutions to complete a 3D model of a structure like the Cyrene Treasury. Other complex architectural components, such as the lionhead rainspouts, were modeled in Autodesk’s Mudbox, software which facilitates the sculpting of organic forms. The file size for each lionhead was larger than the entire Cyrene structure combined. Because of the extraordinary number of polygons associated with the shape of the lionheads and the Ionic door molding, those architectural features were not included in the 3D virtual reconstruction of the treasury that was utilized in the implementation of the prototype immersive gesture-based environment. 3. INTERACTIVE ASHES2ART PLATFORM The fusion of architecturally precise models, such as the current Cyrene Treasury, with a virtual landscape reconstruction and linked data is not groundbreaking. However, emerging gesture based learning © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 25-33 30 ARNE FLATEN et al technologies, such as Microsoft’s Kinect, are changing how we interact with information within a virtual landscape environment (Richards-Rissetto et al. 2013). This newest component to the larger Ashes2Art project was born out of the desire to physically interact with digital technology in a humanistic manner free from extraneous hardware limitations such as 3D glasses, joysticks, keyboards, or virtual reality headsets. In addition, the design and development of the prototype interactive Ashes2Art platform was driven by student interest in leveraging new technologies to bring a more dynamic, humanistic perspective to working with the myriad sources and types of information related to the structures and site of ancient Delphi. Students wanted to go beyond viewing individual models in a limited viewer, and develop an immersive, interactive environment that could provide not only a landscape context for the virtual reconstructions, but also contextualize the digital information that was developed during the 3D modeling process. The idea was the brainchild of two undergraduate students at Coastal Carolina University, who proposed to leverage a consumer videogame interface device, Microsoft’s Kinect controller, and develop a custom interactive exploration environment. This platform would combine the ability to explore 3D virtual reconstructions of the structures and landscape of ancient Delphi with functionality to delve more deeply into the scholarship of individual structures and monuments through embedded multimedia and database links. 3.1 Cyrene Treasury Case Study In order to implement a working prototype of the interactive Ashes2Art platform, a case study 3D model and associated multimedia and scholarly sources were assembled. Since one of the students who initiated the project to design and develop a gesture and audio-based interactive platform had recently completed the research and modeling of the Treasury of the Cyrenes, this structure was chosen as the focus of the prototype demonstration. The students worked extensively with the immersive spatial experience engine (Bergeron 2011) platform to write audio commands, code gesture-base commands, import a basic topography of Delphi, and import a simplified model of the Cyrene Treasury. Ideally, all information from the Fouilles de Delphes on the Cyrene Treasury, including 113 pages of text, 39 plates of drawings and 12 photographic plates (Bousquet and Fomine 1952), would be embedded into the dynamic model and made accessible online, as well as other articles, primary sources, and commentaries. Such a “smart” model would allow access to legacy data consisting of traditional scholarship, a detailed report of our reconstruction methods, and various digital media. The supporting material would then serve as a virtual bibliography to support the work represented in the three dimensional reconstruction, and users would be able to cite this material for academic use. We can read about the Cyrene Treasury and see the engineering expertise of its builders represented in a virtual reconstruction, but having the ability to manipulate scale and perspective enhances the experience significantly. Users want the freedom to remove a column drum with their hands, for example, and rotate it to examine the flutes of the shaft. They want to see how high the triglyphs are by simply looking up. They want the intrigue or interaction of an immersive first-person camera based video game coupled with the research capabilities of a university library. Our template, still in its early stages, attempts to create a gesture-based learning platform with academic research capabilities. 3.2 Developing the platform One of the students’ main goals in developing their prototype was to define a basic set of relatively intuitive Kinect gestures and voice commands to create a more dynamic platform for presenting infor- © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 25-33 IMMERSIVE TECHNOLOGIES AT DELPHI mation about the ancient monuments of Delphi that also fosters collaborative discussions and active participation by students and instructors. The prototype developed for this pilot project accomplished these goals by combining an immersive virtual landscape reconstruction with informational screens and multimedia to delve further into the archaeological and scholarly evidence used to develop each reconstruction (Bergeron 2011). The Ashes2Art interactive platform is built on a common videogame design that utilizes a series of screens that are loaded when the application runs and then accessed as needed through gestures, voice commands or, if needed, keyboard or controller input. Opening screens provide an introduction to the project and give the user (whether student or scholar) information needed to get started with the application. 3.3 Screen Design Following the informational opening screens, the user is given a menu of options, including the choice to enter an immersive virtual landscape representation of ancient Delphi and its reconstructed monuments or a choice to explore interactive 3D architectural views of individual monuments. 31 mands and simple Kinect gestures were added to keyboard and controller navigation input to give users a range of options for exploring the immersive Delphi landscape. The main focus of the platform development, then, was in the design and implementation of the informational screens that would present the wealth of scholarly information and multimedia that explores the Treasury of the Cyrenes and its virtual reconstruction. When a user navigating the virtual Delphi landscape approaches a monument or structure he or she wishes to explore more deeply, a simple voice command to “explore” or a hand gesture to select the structure triggers the loading and display of a main Explore Screen, featuring general information about the structure and a simplified 3D model. This screen is the jumping-off point to additional screens that show links to multimedia and digital scholarly sources, animations and details of parts of the monument’s construction, such as a column (Fig. 5). Figure 5. Interactive exploration screen within custom designed XNA platforms Figure 4. Detail of virtual Delphi landscape development in custom designed XNA platform For the first phase of prototype development, the ancient Delphi landscape reconstruction was not the main focus, and was developed using only simple block models of the major monuments and a generalized terrain (Fig. 4). Voice com- Users can also interact with individual elements of a 3D model through a screen that allows them to grab and manipulate construction elements with Kinect hand gestures (Figure 6). Not only does this provide an interactive educational experience for users, but also demonstrates how the 3D model itself was developed through the © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 25-33 32 ARNE FLATEN et al interpretation of available archaeological and scholarly sources. Figure 6.Interacting with 3D model immersive elements using hand gestures and voice commands (Photo credit: M. Garofalo) Users can move freely between the screens using Kinect gestures such as select to press virtual buttons, voice commands, and even keyboard and controller input if needed. The natural user interface of the interactive Ashes2Art platform also allows multiple users to collaborate and explore the virtual environment together, as the Kinect can accept input from multiple voices and track multiple users’ movements. 4. CONCLUSION This interdisciplinary project has allowed undergraduate students to work closely with faculty mentors in a number of fields, and these students presented the results of their research, including a technical prototype, at the 41st Computer Applications and Quantitative Methods in Archaeology conference in Perth, Australia (Garofalo 2013; Rudolph 2013). They were also chosen to present their work at the 2013 Posters on the Hill event in Washington, DC where they discussed their work with U.S. Senators and Congressmen. Posters on the Hill showcases outstanding undergraduate research at US colleges and universities, hosted by the National Council on Undergraduate Research. In the future, we hope to extend the interactive Ashes 2Art platform to include an integrated search engine that would allow for additional research with a separate display to show internet search results while allowing users to remain within the immersive environment. Future applications might also incorporate social media plugins to promote platform use and product awareness. Another possible feature could be a Wiki platform with a user submission form. We still have much to do. Our ultimate goal is to have archaeometric 3D models of every structure at Delphi placed within a GIS-based, topographically accurate virtual landscape with navigation and interaction via an immersive, gesturebased and voice-controlled environment. This platform, whose viability has been demonstrated by the Cyrene Treasury prototype, will give users access to vast quantities of metadata and multimedia with a simple pinch. Or a swipe. Or a word. At the same time that we continue work on this very large environment, we are also designing an app for smartphones and tablets to navigate the models intuitively. ACKNOWLEDGEMENTS We would like to thank the Hellenic Ministry of Culture, the National Endowment for the Humanities USA, the American School for Classical Studies in Athens, and the Delphi Archaeological Museum. The authors would also like to thank Professor Dr. Ioannis Liritzis, University of the Aegean, for organizing and supervising VAMCT13, the conference where this material was presented, and for supervising the peer review of the present journal. REFERENCES Bergeron, S. J. (2011) Engaging the Virtual Landscape: toward an experiential approach to exploring place through a spatial experience engine. PhD Dissertation, Department of Geology and Geography, West Virginia University, Morgantown, WV. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 25-33 IMMERSIVE TECHNOLOGIES AT DELPHI 33 Bousquet, J. and Fomine, Y. (1952) Le Trésor de Cyrène. École Francaise d’Athènes, Fouilles de Delphes, tome II, topographie et architecture. Paris: E. De Boccard. Cook, R. M. (1954) Review of Le Trésor de Cyrène. The Classical Review, New Series, vol. 4, no. 2 ,177. Dinsmoor, W. B. (1957) Review of Le Trésor de Cyrène. American Journal of Archaeology, vol. 61, no. 4, 402-411. Flaten, A. R. (2009) The Ashes2Art Project: Digital Models of Fourth-Century BCE Delphi, Greece. Visual Resources: An International Journal of Documentation. Special Issue. Digital Crossroads: New Directions in 3D Architectural Modeling in the Humanities, vol. 25, no. 4, 345-362. Garofalo, M. (2013) A Case Study: architectural accuracy, virtual modeling, and gesturebased learning with the Cyrene Treasury at Delphi, Greece. 41st Computer Applications and Quantitative Methods in Archaeology Conference, Perth, Australia, March 24-28. Gill, A. A. (2009) Digitizing the Past: charting new courses in the modeling of virtual landscapes. Visual Resources: An International Journal of Documentation. Special Issue. Digital Crossroads: New Directions in 3D Architectural Modeling in the Humanities, vol. 25, no. 4, 313-332. Gill, A. A. and A. R. Flaten (2008) Digital Delphi: the 3D Virtual Reconstruction of the Hellenistic Plunge Bath at Delphi. In The Digital Heritage: Proceedings of the 14th International Conference on Virtual Systems and Multimedia, M. Ionnides, A. Addison, A. Georgopoulos, L. Kalisperis (eds). Archeolingua, Hungary, 427-30. Harris, T. M., S.J. Bergeron, and L. J. Rouse. (2011) Humanities GIS: Place, Spatial Storytelling and Immersive Visualization in the Humanities. In GeoHumanities: Art, History, Text at the Edge of Place, M. Dear, J. Ketchum, S. Luria, and D. Richardson (eds.) Routledge, New York, 226-240. Haspels, C. H. E. (1953) Review of Le Trésor de Cyrène. Mnemosyne, Fourth Series, vol. 6, Fasc. 3, 242-243. Laroche, D. (1988) "L'emplacement du trésor de Cyrene," Bulletin de correspondance hellénique vol. 112, no. 1, 291-305. Partida, E. C. (2000) The Treasuries at Delphi: An Architectural Study. Studies in Mediterranean Archaeology and Literature, no. 160, Paul Astroms Forlag, Sweden. Plommer, W. H. (1954) Review of Le Trésor de Cyrène. Journal of Hellenic Studies, vol. 74, 223-224. Richards-Rissetto, H., Robertsson, J., von Schwerin, J., Remondino, F., Agugiaro, G., and Girardi, G. (2013) Geospatial Virtual Heritage: A Gesture-based 3D GIS to Engage the Public with Ancient Maya Archaeology. Proceedings of Computer Applications and Quantitative Methods in Archaeology, Southampton, United Kingdom, March 26-29. Rowe, A. et al. (1956) Cyrenaican Expedition of the University of Manchester 1952. Manchester University Press, United Kingdom. Rudolph, C. B. (2013) “Kinecting” cultural heritage sites with immersive audio and gesture-based learning environments. 41st Computer Applications and Quantitative Methods in Archaeology Conference, Perth, Australia, March 24-28. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 25-33 Mediterranean Archaeology and Archaeometry, Vol. 14, No 4, pp. 35-44 Copyright © 2014 MAA Printed in Greece. All rights reserved. METHODOLOGY TO CREATE 3D MODELS FOR AUGMENTED REALITY APPLICATIONS USING SCANNED POINT CLOUDS Radu Comes1, Călin Neamțu1, Zsolt Buna1, Ionuț Badiu 1, Paul Pupeză2 Department of Design Engineering and Robotics, Technical University of Cluj-Napoca, Muncii Avenue, no. 103-105, 400641, Romania 2 National History Museum of Transylvania from Cluj-Napoca, C. Daicoviciu St, no. 2, 400020, Romania 1 Received: 30/11/2013 Accepted: 05/08/2014 Corresponding author: Radu Comes (comesradu@gmail.com) ABSTRACT Precise digital documentation of cultural heritage assets is essential for its preservation and protection. This documentation increases the efficiency of scientific studies that are being carried out during the restoration and renovation process. Precise digital documentation makes use of different laser scanners technologies. 3D scanning devices usually provide a large amount of point clouds, which require long post-processing times and large storage space. This paper presents a methodology to obtain simplified 3D models designed to remove redundant points and maintain only representative points, preserving the 3D model aspect and allowing the 3D models to be implemented in different augmented reality application on mobile devices. The 3D mesh optimization methods that have been analyzed and compared are dedicated 3D mesh optimization software (CATIA and Geomagic Studio), open source tools such (Meshlab) and a numerical computing environment (MATLAB). The methodology proposes a split step that can be applied to both assemblies and reconstructed objects. In this step the 3D scan is divided into components (for assemblies) or original parts/restored parts (for restored cultural heritage assets). The efficiency and robustness are demonstrated using different 3D scanned Dacian artifacts. KEYWORDS: laser scanning, surface reconstruction, geometric deviation, point clouds 36 RADU COMES et al 1. INTRODUCTION 3D imaging techniques, multimedia applications, virtual reality and augmented reality applications become increasingly used in museums and galleries around the world covering a wide range of areas starting with museums dedicated to nature and natural history (Schultz 2013), human science, religious (Gkion, Patoli et al. 2011), etc. Due to the wide availability of 3D scanners and other data acquisition systems, many 3D models are obtained through the scanning process of actual objects from the museum collections. The discrete point set obtained is then processed to a continuous surface representation such as spline, volumetric or polygonal. The increasing availability of low-cost 3D acquisition devices has resulted in the widespread dissemination of 3D point clouds of sampled real-world objects. As a consequence, surface reconstruction from acquired 3D point clouds has become an important problem in computer graphics and computer vision (Seversky, Berger et al. 2011) Many researches (Bruno, Bruno et al. 2010, Rua and Alvito 2011), create detailed models of historical artifacts and then create simplified versions for various virtual reality and augmented reality applications. Creating simplified versions of 3D models is a critical task because 3D model simplification leads to a decrease in quality. As shown in (Meftah, Roquel et al. 2010), reducing the amount of data that is used to generate a 3D model implicitly generates errors and deviations of the shape that leads to “deformed” 3D models thus reducing its quality. Level of Detail (LOD) is an indicator that can be taken as an indicator of the quality of the 3D model. There are several approaches when creating virtual reality and augmented reality applications for museums and exhibitions that focuses on different domains such as: mixed-reality learning (Wu, Lee et al. 2013), augmented reality in a public space (Barry, Thomas et al. 2012), interactive experiences through a 3D reconstruction (Gkion, Patoli et al. 2011), virtual exhibitions on multitouch devices, developing serious games for cultural heritage (Anderson, McLoughlin et al. 2010), low-cost creation of a 3D interactive museum exhibition (Monaghan, O'Sullivan et al. 2011), evaluating interaction in an online virtual archaeology site, etc. The authors propose a methodology that can provide optimize 3D models for augmented reality applications. The 3D models are obtained using laser scanning and thus a large amount of data is acquired. This paper focuses on the creation of 3D simplified content for mobile devices and is focused on Dacian cultural heritage assets. 2. PROPOSED METHODOLOGY The methodology proposed is illustrated in Figure 1. Figure 1 Methodology used to create simplified 3D laser scanned models for mobile augmented reality applications Technologies and methodologies for the digitization of cultural heritage assets are wide-ranging. Our approach has distinct © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 35-44 METHODOLOGY TO CREATE 3D MODELS FOR AUGMENTED REALITY APPLICATIONS characteristics and makes use of laser scanners and mesh simplification to implement the 3D scanned model on mobile devices. 2.1 Point Cloud The first phase of the methodology consists in the 3D scanning and the processing of the acquired data. The objects are scanned from various positions, in order to acquire the full shape. For pots and other vessels, the internal hollow part is also scanned. There are many methods to acquire 3D data. Each method uses different mechanisms to interact with the surface or the volume of a real object. During the 3D scanning process, there are certain problems that can appear. Most of them are related to the optical limits of certain laser scanners. Laser scanning give notoriously poor results when shiny surfaces are scanned. These surfaces tend to absorb the light beams, modifying the light environment and in some cases coating the artifacts with a thin sprayed white matte layer improves the scanning. Using current scanning technologies uniform points clouds (distance between points is constant Figure 2.a) and nonuniform (distance between points is variable Figure 2.b), can be obtained. The proposed methodology can be used for both types on point clouds. 37 to coordinate measuring machine (Figure 3.c). Contact scanning can be used also, but only if the contact between the ruby ball of the touching probe and the artifact surface does not damage the artifact. (Figure 3.d) Figure 3 a - hand held laser scanner (ViuScan) b – laser scanner Zephyr Z-25 attached to portable a CMM (Stinger II) c – laser scanner attached to CMM(Scope-Check) d – contact scanning (Cyclone series 2) The main advantage of 3D laser scanning is its non-invasive mechanism and high speed/good resolution of data record. These scanning devices usually generate dense cloud points with millions of points which require a large storage space and long post-processing time. The point cloud density is directly proportional with the accuracy of the 3D scan. Many of the scanners available on the market provide real time automatic generation of the mesh. The first step of the proposed methodology also involves the processing of the point cloud. This involves removing the points that represent the element that was used as a support for the scanned artifact or noise (Figure 4). Figure 2 Point clouds type: a – uniform (ceramic vessel) and b – non-uniform (anvil) As shown in (Savio, De Chiffre et al. 2007, Morovic and Pokorny 2012, OK Rahmat, Ng et al. 2012) scanning small and medium size artifacts can be done with hand held scanners (Figure 3.a), scanners attached to portable coordinate measuring machine (Figure 3.b) or scanners attached Figure 4 Removing points from Bendis scanned model © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 35-44 38 RADU COMES et al Points removal can be made in CAD software using specific tools or software that allow mathematical calculation and removal of points based on the interpretation of XYZ coordinates or RGB attributes. 2.2 Filter The objective of this step is to reduce the number of points so that further processing can be done easier. The higher the number of points acquired is the higher the number of polygons will be. A higher number of polygons will approximate better the real surface of the scanned object. There are several types of filters (Cignoni, Montani et al. 1998) developed for processing point clouds: Gaussian, uniform, median, morphological, binary, derivatives, etc. These types are being used in CAD software or individual software (e.g. Metro (Cignoni, Rocchini et al. 1998)). Using adaptive filtering (Figure 5) the areas with a high amount of details will have a denser points cloud mesh, while flat areas or smooth curves have a much lower point density. Figure 5 Example of filtering methods 2.3 Mesh A mesh (M) is a discrete representation of geometric model in terms of its geometry G, topology T, and associated attributes A. Were: G - geometry - nodal coordinates T - topology - element types, adjacency relationships A - attributes - color, boundary conditions, etc. Geometrically a mesh can be classified by the shape of the geometric elements that are used to generate the mesh: triangles, polygons, quadrangles, polyhedral, tetrahedrons or hexahedrons. AR applications commonly use polygonal and triangle meshes to reduce the number of vertices that has to be processed. The mesh can be generated using the acquired point cloud or there are some scanning software solutions that can generate the mesh using the scanning software (VX elements, Figure 6). Figure 6 Mesh generation in VX elements Bernardini et al. (Bernardini, Mittleman et al. 1999) developed a Ball Pivoting Algorithm (BPA) for surface reconstruction from a given point cloud. The principle is simple: three points form a triangle if a ball of a user-specified radius touches them without containing any other point. Therefore, starting with a seed triangle, the ball pivots around an edge until it touches another point, forming another triangle. After the mesh has been generated it requires addition processing so that the elements that are manifold, self-intersecting, small and thin can be removed. Besides these operations that are predecessor to the surface generation process, other operations must be conducted such as: smoothing sharp edges and filling holes. 2.4 Mesh Simplification Most of the simplification methods for 3D models are focuses on polygonal models because all others representations can be converted to a polygonal one. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 35-44 METHODOLOGY TO CREATE 3D MODELS FOR AUGMENTED REALITY APPLICATIONS Mesh simplification aims to reduce the number of polygons and to keep the shape as close to the initial shape, as the number of polygons is reduced the processing time decreases. In some cases, better results can be obtained if the point clouds are heavily filtered before the mesh is generated. The simplification of point-based 3D models should satisfy several requirements: (i) the re-sampling of the points from the large point set should be performed efficiently; (ii) the desired reduced number of output samples should be controlled; (iii) the distribution of the output samples can be controlled; (iv) the difference between the original surface and the corresponding surface after resampling should be minimized (Yu, Wong et al. 2010). 2.5 Split component / part In some cases the object cannot be disassembled to allow individual scanning of each component. For artifacts that cannot be disassembled the object can be scanned as a whole and then split in the virtual environment. In these situations the assembly components can be extracted using geometrical operations. After the components have been extracted they can be reassembled in the virtual environment. These should be reconstructed using CAD software so that the component can be completely reconstructed. The figure below presents the case of a Dacian iron plier that has been preserved in the open position, without the possibility to close the jaws. Using CAD software the scanned model was divided into three parts as shown in Figure 7. 39 Figure 7 Dacian iron plier virtual restoration 2.6 Quality check The objective of this phase is to determine the maximum deviation between the simplified 3D models and the scanning results. Different tools can be used such as: Deviation Analysis available in almost any CAD solution, Hausdorff distance, or independent solutions tools developed in mathematical computation software (MATLAB, Mathematica, etc.) Figure 8 Deviation analyses using four different software Figure 8 displays four deviations analysis, two were performed in CAD solutions (CATIA V5R21 and Geomagic Studio 2012) and the other two were conducted in MATLAB and Meshlab. The 3D model is a scanned Dacian hammer, the point cloud obtained using the laser scanning had 140,780 polygons and 70,394 vertices, while the simplified model had 23,463 polygons and only 11,733 vertices. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 35-44 40 RADU COMES et al The geometrical differences between the four methods are different but all the values are in the range of 0.003 mm for the minimum deviation and 0.002 mm for the maximum deviation. The main advantage of CATIA is that it provides real time interaction with the analyzed object, using the mouse cursor specific points and their deviation values will be overlapped over the 3D model as shown in Figure 9. Figure 9 Interactive deviation analyses from CATIA Figure 10 presents the deviation analyses using the Hausdorff distance tool within Meshlab. The values are similar to the one obtained in CATIA. Figure 10 Hausdorff distance analysis within Meshlab Figure 11 presents the deviation analyses using MATLAB, the values are similar to the ones obtained in the other software. Figure 11 Deviation analysis using MATLAB Since the deviations are so small, the authors consider that the 3D simplified model is suited to be used in augmented reality application. But if the artifact is intended to be used for different research activities, the 3D model that hasn’t been optimized should be used. 2.7 Texture Texture is the operation that offers visual esthetic skin to 3D models so that the virtual object will be similar to the real object. Texturing can be done in several ways, both in CAD software solutions and in software solutions that are used in the gaming and animation industry (3ds max, Maya, Blender, etc.). This operation can be done in two distinct moments of the methodology: right after the quality check or directly in the software to create the AR application, if the application support texturing. 2.8 3D conversion The most commonly used format used in the AR applications are *.obj, *.vrml, *.3dxml and .dae, since these formats can contains additional information, such as texture. There are scanning solutions that are able to provide *.obj file format directly, but the most common scanners offer standard CAD formats such as *.stl, *.igs, *.stp, *.ascii. In most cases mesh processing is done in other formats than those mentioned above, because of this 3D conversion is required to be able to use them in AR applications. 3D conversion can be done using specialized software (Deep exploration), directly in applications that were used to process the scan mesh, web converting solutions, or solutions developed in different programming software. During the conversion of 3D scanned operation, the geometry may suffer some changes. In these cases additional operations are required to ensure the correct conversion of the 3D models. One of these operations involves the generation of a surface created on top of the initial pro- © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 35-44 METHODOLOGY TO CREATE 3D MODELS FOR AUGMENTED REALITY APPLICATIONS cessed mesh. This surface will be exported instead of the processed mesh. The alternative to this solution is to generate solid bodies and then convert them to avoid geometry changes. 41 The laser scanning generated a point cloud of about 340.000 points. 2.9 Animation 3D animated models provide a better visual experience compared to a static 3D model. In the case of simple objects, automatic rotation around z axis provides an overview of the object and eliminates the need for manual manipulation. In the case of complex assemblies or components, animation can provide additional information to the user. 2.10 Metadata Metadata represents all the additional data that will be added to the 3D model to provide a more enjoyable 3D experience to the user. These can be text file, audio files, 2D images or video files. These metadata are intended to provide additional information when the user requests them, while viewing the 3D model in the augmented reality environment. The metadata helps users to better understand the virtual artifacts. 2.11 Testing on mobile device The authors used three different mobile devices (two smartphones and one tablet) to test the augmented reality application. The VaD AR application has been created using Metaio Creator and it contains 15 different 3D scanned models. VaD AR uses only 2D image tracking at the moment, but the authors will extend the application to use different 3D objects and environment tracking system such as SLAM instant tracking and visual search. Figure 12 Dacian pliers a- real photo b- point cloud The first hypothesis was to reduce the number of points using filtration and then generate the mesh. The second one was to generate the mesh using all the points and then reduce it gradually using different filters and software solutions. The table bellows (Table 1) provides the reports about the number of points and polygons obtained by the two methods. Level of detail 90% 80% 70% 60% 50% 40% 30% 20% 10% Table 1 Simplified 3D models Cloud point Mesh Vertices count Polygons count Vertices count Polygons count 157,013 135,652 118,721 101,990 84,866 68,007 51,088 33,972 16,969 325,707 282,458 247,152 211,843 176,537 141,232 105,923 70,616 35,308 156,758 135,819 118,844 101,867 84,892 67,916 50,938 33,961 16,983 313,180 271,595 237,647 203,696 169,748 135,799 101,849 67,900 33,950 5. RESULTS In order to validate the methodology a case study was conducted. A scanned Dacian pliers (Figure 12) is illustrated below, the artifact was scanned using a laser scanner that provides an accuracy of 15 μm. Figure 13 Information regarding the scanned Dacian pliers © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 35-44 42 RADU COMES et al In order to check the quality of the 3D simplified models (Figure 13), the simplified meshes were compared with the initial mesh that was generated using the initial data acquired by the laser scanner. Level of detail Table 2 Simplified 3D models MATLAB Min 90% 80% 70% 60% 50% 40% 30% 20% 10% and 3,396 polygons. Because this object has been simplified too much, many details are missing. -0.0010 -0.0028 -0.0040 -0.0097 -0.0133 -0.0172 -0.0301 -0.0477 -0.0972 Max 0.0011 0.0021 0.0044 0.0073 0.0118 0.0243 0.0340 0.0554 0.1404 CATIA V5 Min Max -0.0011 -0.0021 -0.0043 -0.0071 -0.0114 -0.0234 -0.0327 -0.0533 -0.1005 0.0010 0.0027 0.0039 0.0094 0.0128 0.0166 0.0290 0.0459 0.0935 The results from CATIA V5 R21 and MATLAB are presented in the table 2. The values obtained are similar. But if we compare this deviations with the ones presented for the digitized hammer in Figure 9, we see that the deviations are constantly increasing very rapidly along with the 3D model simplification. Figure 14 Rivet visual deviations The highest deviations are located near the round rivet that keeps the two pliers jaws together. The difference between the original mesh and the 10% simplified model can be seen in Figure 14. The rivet roundness has been deformed. A balance between the degree of fidelity of a 3D model details and file size must be made, so that the 3D model can be loaded fast on mobile devices such as smartphones or tablets. Even more simplified 3D models have been created such as the LOD 1% (Figure 15). This 3D model has only 1,701 vertices Figure 15 Optimized 3D models surfaces The table below presents the file size in KB for three common 3D file formats, the stereolithography (*.stl), the wavefront technologies (*.obj) and the 3DVIA (*.3dxml). These files describe the surface geometry of a 3D model. The *.stl file format has no texture, while *.obj and *.3dxml can support texture or other common CAD model attributes. (Table 3) The most common 3D file format for mobile devices is the *.obj file format. Recently augmented reality applications have started to use *.fbx files, the main advantage of this file format is that the animations are embedded within the same file. Table 3 File format and file size in KB Level of *.STL *.OBJ *.3DXML Detail LOD 100% LOD 90% LOD 80% LOD 70% LOD 60% LOD 50% LOD 40% LOD 30% LOD 20% LOD 10% 16,577 14,929 13,262 11,604 9,947 8,289 6,631 4,974 3,316 1,658 34,256 30,361 26,774 23,189 19,606 16,272 12,977 9,683 6,389 3,096 4,862 4,394 3,921 3,449 2,970 2,493 2,010 1,528 1,043 558 © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 35-44 METHODOLOGY TO CREATE 3D MODELS FOR AUGMENTED REALITY APPLICATIONS The simplified 3D models have been tested on three applications. Two applications are free to use from Google play store: 3D AR and AndAR. The other application is VaD AR (Figure 16). 43 mobile devices can load the proper models and overlap them correctly. In order to make the application viable for a much higher amount of 3D models, the management system within the application requires more fine-tuning. 5. CONCLUSIONS Figure 16 VaD AR augmented reality application The table below (Table 4) presents the initial loading time (in seconds) of the 3D simplified LOD 10% model of the Dacian pliers using different AR applications. Table 4 LOD 10% 3D model of the Dacian pliers average loading time (seconds) on mobile devices Mobile device Asus TF300T Samsung Galaxy Note II N7100 Samsung I9300 Galaxy S III 3D AR And AR VaD AR 14 15 12 12 13 11 15 16 15 After the models are loaded the user can focus the camera on different markers and the application will display the correct model instantly with no loading time. A total of 15 simplified 3D models have been added to the VaD AR application. The Creating simplified 3D models from data acquired with different laser scanning techniques is not a simple task. The simplification method needs to be carefully analyzed so that the simplified model deviation analyses are not high. The methodology presented in this paper can be used to obtain simplified 3D models that can be loaded into commercial and custom made AR applications. Reducing the polygons number of a 3D model entails a lower level of detail, and in some cases a loss of detail of the scanned object. Polygon reduction can be done in different CAD software application or in custom applications created specifically for this purpose. After testing two CAD software solutions (CATIA V5, Geomagic Studio 2012) and a software solution that uses a Gaussian reduction algorithm implemented in MATLAB, no significant differences have been recorded between the deviation values of the 3D optimized models. The VaD AR application with 15 simplified 3D models and 15 different markers has been presented with the occasion of the Night of Museums cultural event in ClujNapoca. The public enjoyed interacting with 3D scanned artifacts using augmented reality. 6. ACKNOWLEDGMENTS This paper is supported by the Sectoral Operational Programme Human Resources Development POSDRU /159/1.5/S/ 137516 financed from the European Social Fund and by the Romanian Government. REFERENCES "Metaio Creator" from http://www.metaio.com/creator/. Anderson, E., L. McLoughlin, F. Liarokapis, C. Peters, P. Petridis and S. de Freitas (2010) Developing serious games for cultural heritage: a state-of-the-art review. Virtual Reality 14(4): 255-275. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 35-44 44 RADU COMES et al Barry, A., G. Thomas, P. Debenham and J. Trout (2012) Augmented Reality in a Public Space: The Natural History Museum, London. Computer 45(7): 42-47. Bernardini, F., J. Mittleman, H. Rushmeier, C. Silva and G. Taubin (1999) The ballpivoting algorithm for surface reconstruction. IEEE Transactions on Visualization and Computer Graphics 5(4): 349-359. Bruno, F., S. Bruno, G. De Sensi, M.-L. Luchi, S. Mancuso and M. Muzzupappa (2010) From 3D reconstruction to virtual reality: A complete methodology for digital archaeological exhibition. Journal of Cultural Heritage 11(1): 42-49. Cignoni, P., C. Montani and R. Scopigno (1998) A comparison of mesh simplification algorithms. Computers and Graphics, 22(1): 37-54. Cignoni, P., C. Rocchini and R. Scopigno (1998) Metro: Measuring Error on Simplified Surfaces. Computer Graphics Forum 17(2): 167-174. Gkion, M., Z. Patoli and M. White (2011) Museum interactive experiences through a 3D reconstruction of the Church of Santa Chiara. Proceeding (744) Intelligent Systems and Control / 742: Computational Bioscience - 2011. Meftah, A., A. Roquel, F. Payan and M. Antonini (2010) Measuring errors for massive triangle meshes, 2010 IEEE International Workshop on: Multimedia Signal Processing (MMSP): 379-383. Monaghan, D., J. O'Sullivan, N. E. O'Connor, B. Kelly, O. Kazmierczak and L. Comer (2011) Low-cost creation of a 3D interactive museum exhibition. Proceedings of the 19th ACM international conference on Multimedia. Scottsdale, Arizona, USA, ACM: 823-824. Morovic, L. and P. Pokorny, Optical 3D Scanning of Small Parts, in Automation Equipment and Systems, Pts 1-4, W.Z. Chen, et al., Editors. 2012, Trans Tech Publications Ltd: Stafa-Zurich. p. 2269-2273. OK Rahmat, R. W., S. B. Ng and K. Sangaralingam (2012) Complex Shape Measurement Using 3D Scanner. Jurnal Teknologi 45(1): 97–112. Rua, H. and P. Alvito (2011) Living the past: 3D models, virtual reality and game engines as tools for supporting archaeology and the reconstruction of cultural heritage – the case-study of the Roman villa of Casal de Freiria. Journal of Archaeological Science 38(12): 3296-3308. Savio, E., L. De Chiffre and R. Schmitt (2007) Metrology of freeform shaped parts. CIRP Annals - Manufacturing Technology 56(2): 810-835. Schultz, M. K. (2013) A case study on the appropriateness of using quick response (QR) codes in libraries and museums. Library & Information Science Research 35(3): 207215. Seversky, L. M., M. S. Berger and L. Yin (2011) Harmonic point cloud orientation. Computers & Graphics 35(3): 492-499. Wu, H.-K., S. W.-Y. Lee, H.-Y. Chang and J.-C. Liang (2013) Current status, opportunities and challenges of augmented reality in education. Computers & Education 62(0): 41-49. Yu, Z., H.-S. Wong, H. Peng and Q. Ma (2010) ASM: An adaptive simplification method for 3D point-based models. Computer-Aided Design 42(7): 598-612. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 35-44 Mediterranean Archaeology and Archaeometry, Vol. 14, No 4, pp. 45-54 Copyright © 2014 MAA Printed in Greece. All rights reserved. VISUALIZING THE HISTORY AND ANALYZING THE SCULPTURAL DECORATION OF THE TEMPLE OF ZEUS AT OLYMPIA IN VIRTUAL REALITY András Patay-Horváth University Eötvös Loránd, Institute for Ancient History (& Hungarian Academy of Sciences, Archaeological Institute), Egyetem tér 1-3. Budapest, 1053 Hungary Received: 27/01/2014 Accepted: 26/06/2014 Corresponding author: A. Patay-Horváth (pathorv@gmail.com) ABSTRACT The project presented here started approximately four years ago and concerns the main temple of Olympia, a UNESCO world heritage site, which is visited by thousands of tourists nearly every day. Although Olympia is familiar to everybody and its monuments have been well-researched for more than a century, there are still many puzzles related to its history and remains. A new interpretation of the east pediment of the temple and the ensuing debate caused the reopening of the issue of the reconstruction. The historical setting of the temple-building was also reconsidered and led to a detailed study and reconstruction of the architecture as well. All these investigations made extensive use of digital technologies and are presented here as a case study for applying virtual reality to old problems of classical archaeology. The digitization of the extant fragments and a three-year project enabled the production of a virtual 3D reconstruction of the east pediment of the classical temple of Zeus. In addition, the Doric temple itself as well as the famous cult statue made by Pheidias were also reconstructed virtually, making thus the visualization of the long and complicated history of the entire monument possible. The model is highly flexible and can thus be adapted to illustrate and to test different scholarly hypotheses concerning some details, e.g. the arrangement of the central group of the east pediment or the effects of different lighting conditions. It also allows the nonspecialist user to manipulate the individual pieces of sculpture, to familiarize him- or herself with their original appearance and position on the building and finally to observe minor details and to learn more about the problems involved in reconstructing ancient works of art. A short video-summary and a CD ROM have been published, both of which can be used for different purposes and audiences. KEYWORDS: 3D modelling, virtual reconstruction, 5th century BC, Greece, architecture, pedimental sculpture, Pheidias. 46 ANDRÁS PATAY-HORVÁTH 1. INTRODUCTION The project presented here started approximately four years ago and concerns the main temple of Olympia. Although Olympia is familiar to everybody and its monuments have been well-researched for more than a century, there are still many puzzles related to its history and remains. The present project started from a recent controversy surrounding the interpretation of the east pediment. 1 2. HISTORICAL BACKGROUND The temple (Figure 1) was built in the early classical period, ca. 475-455 BC. 2 At the time of its construction, it was the largest temple in mainland Greece, and it has remained the largest ancient temple on the Peloponnese. Given the large size of the building itself, the sculptural decoration was well over lifesize and was made of white marble. A large number of fragments survive in a fairly good condition. They are depicted in practically every handbook on Greek art, because nowadays they are considered to be one of the most important and most magnificent works of ancient Greek art. They are contemporary with the building itself. For ten to fifteen years after the completion of the building and its outside decoration, the temple seems to have been empty, i.e. without any cult statue. The famous gold-ivory statue of the seated Zeus by Phidias, one of the seven wonders of the ancient world, was only erected in a second building phase ca. 440430 BC. 3 Afterwards, there was a tremendous earthquake in 373BC which caused considerable damage and several subsequent rebuilding and restoration episodes and also later additions, like the Patay-Horváth 2008. For earlier reports see Patay-Horváth 2011a; 2011b; 2011c; 2012; 2013a. 2 For the temple in general and for the date see e.g. Dinsmoor 1950, 151-153; Mallwitz 1972, 211-234; Lawrence 1983, 184-185; Gruben 2001, 56-62; Hellmann 2002, 124; Lippolis et al. 2007, 385-389, 655-657. For the historical circumstances Patay-Horvαth 2013b. 3 Schiering 1991; Strocka 2004, 228. 1 21 golden shields hung up after 146BC on the cornice. 4 Figure 1 Reconstruction of the temple as it was seen by visitors in 2nd century AD. 3. THE RECONSTRUCTION OF THE TEMPLE IN VIRTUAL REALITY The long and complicated story of the building can only be narrated or vizualized with a series of reconstruction drawings. A flexible digital model, like the one produced during our project (Figure 2), is much more suitable for this purpose and offers additional features, which would be hardly feasibly with any traditional model. Figure 2 Uncolored virtual 3D reconstruction of the temple as completed around 450 BC. The best illustration of the possibilities is the simulation of the lighting conditions in the interior of the temple. A recent attempt without the help of a digital model envisaged two possibilities: sunlight comes either only through the door (Figure 3), or through a hole in the roof (Figure 4). 5 For the 4th century BC see esp. Grunauer 1981. For all the renovations Hennemeyer 2010. For the shields dedicated by Mummius Paus. 5, 10,5. 5 Hennemeyer 2011, 101–104. 4 © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 45-54 THE TEMPLE OF ZEUS AT OLYMPIA IN VIRTUAL REALITY Figure 3 The cult statue in its architectural setting (above: cross section of the temple after Hennemeyer 2011 fig. 1; below: virtual 3D reconstruction). Illumination through the doors. 47 god would remain dark. The second possibility would be better than the first, but the actual remains of the temple do not support this idea (there is nothing to suggest a hole in the roof). The placement and the measurements of the pool in front of the statue are absolutely certain but as the digital model clearly shows, it can not effectively be used to illuminate the head of the statue. The best illumination would be by direct light from above, and this possibility is perfectly feasible, if we suppose an open ceiling, instead of a hole in the roof. Rooftiles are made of white marble and are therefore transparent and most probably they were employed exactly in order to achieve this lighting effect. 6 (Figure 5) Figure 5 Illumination of the cult statue through translucent marble tiles and an open ceiling. Figure 4 The cult statue in its architectural setting (above: cross section of the temple after Hennemeyer 2011 fig. 2; below: virtual 3D reconstruction). Illumination assuming a hole in the roof. In either case, the light would fall on a shallow pool filled with olive oil in front of the statue and could illuminate just the footstool or the lower part of the statue. Both arrangements would not be particularly satisfactory, because the upper part and especially the head of the seated Curiously enough, this possibility was not considered earlier, and in any case it was impossible to test it without a digital model, but is actually favoured even by Hennemeyer, who opted for the hole in the roof a few years ago. 7 The virtual 3D model can thus be used not just for vizualizing earlier research results and to test earlier hypothesis, but also to improve our understanding of the monument. This is actually even more appropriate in the case of the pedimental sculptures. Ohnesorg 1993, 118–119 with Plin. NatHist 36,46. 7 Hennemeyer 2012, 123. 6 © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 45-54 48 ANDRÁS PATAY-HORVÁTH 4. THE RECONSTRUCTION OF THE EAST PEDIMENT 4.1 The problem of the central group The surviving fragments are substantial and numerous enough to enable a fairly reliable reconstruction. This was done at the end of the 19th century and has duely received general acceptance till today. 8 The only detail which is still controversial is seemingly a minor one, but is actually crucially important for the interpretation and concerns the arrangement of the five central figures of the east pediment; it has been continously debated among archaeologists and art historians since the discovery of the fragments more than a century ago. 9 The basic problem is that the fragments themselves (Figure 6) can be arranged in four substantially different ways (Figure 7) and there are no obvious clues for choosing the most probable one. There is a fairly detailed description of the group by Pausanias, who saw it in the 2nd cent. AD, but his text (Description of Greece, book V, ch. 10, 6-7) is not conclusive regarding the precise arrangement of the figures. The locations of the finds are not unequivocal either, since the pieces were scattered around the temple by an earthquake in the 6th cent. AD and the fragments were subsequently reused in medieval buildings. "B" Open arrangement "A" "A" Closed arrangement "B" Figure 6 Fragments of the central group of the east pediment, as displayed in the Archaeological Museum of Olympia today. Treu 1897; most recently: Heilmeyer et al. 2012. 9 For an overview of the debate cf. Herrmann 1987 and Patay-Horváth 2008. 8 Figure 7 Schematic reconstruction drawings showing every conceivable arrangement of the five central figures. Different colours highlight the differences of the four versions. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 45-54 THE TEMPLE OF ZEUS AT OLYMPIA IN VIRTUAL REALITY 4.2 Earlier reconstructions Most often the reconstructions were presented in simple drawings, ignoring the three-dimensional form of the statues or in miniature plaster models. These miniature models are beautifully coloured, but they are actually not quite accurate. Produced immediately after the discovery of the fragments, they represent only the first attempt and not a final, elaborate version of the reconstruction. 10 Actually, they had been replaced already by the end of the 19th century, with life-size models, which were made by using the plaster casts of the fragments and by restoring the missing parts in plaster as well. The most important result, based on long experimentation with them was summarized by G. Treu as follows: „Sodann ergeben sich bei einer Aufstellung von K* neben Pelops unüberwindliche räumliche Schwierigkeiten. Es wird, wenn man Pelops die richtige, durch die Rückendübel angezeigte Dreiviertelsdrehung zur Ecke hin giebt, unmöglich, seinen Speer an dem schleierfassenden linken Arm von K* vorbei zu bringen. Davon überzeugt ein Versuch mit den Abgüssen in dem richtig gebauten Rahmen ohne weiteres.” 11 Treu stated thus explicitly, that figures G and K can not be placed next to each other, because their arms would come into contact. Obviously, he was absolutely convinced, that this arrangement (Open Type „A”) is physically impossible and invited everybody to verify this statement with the life-size plaster models. This has been done by various scholars following him, and no one questioned this observation. 12 But afterwards, the results of the early experiments were totally ignored: the models were not used for experimentation after World War II and they are no longer mentioned in recent publications, and no one has attempted to verify or to refute this Patay-Horváth 2012. Treu 1897, 120. 12 Studniczka 1923, Bulle 1939. 10 11 49 result. 13 Instead, a great number of studies, and two complete monographs were published on the east pediment, but no-one was able to present a fully satisfactory reconstruction. 14 It is symptomatic that a pair of renowned Greek-English authors presented two completely different reconstructions side by side in the same volume. 15 After a while it seemed that all conceivable arguments had been formulated and no approach proved to be entirely convincing; thus archaeologists grew tired of a seemingly unproductive debate and they gradually agreed on a reconstruction, which was proposed by a few famous scholars. 16 But in this way, an absurd situation emerged: nowadays, the most widely accepted reconstruction is precisely the one (Open Type”A”), which was declared to be technically impossible by Treu. Obviously, this would not present a problem, if his results had been thoroughly tested and clearly refuted, i.e. if anyone had shown that Treu had experimented with poorly-restored models or had come to incorrect conclusions for some other reason. Instead, everyone has ignored his arguments and his results without any discussion. Apparently nobody realized that the best evidence for the benefit of experimenting with life-size models is provided by Treu himself, who had advocated the arrangement widely accepted today, while he only had the miniature models at his disposal, but later his experiences with the life-size models made him change his mind. In spite of the widespread acceptance of this particular arrangement several scholars expressed their doubts and reservations and proposed either other solutions or emphasized that the problem is still open to As far as I know, the models and the results achieved by experimenting with them is mentioned only by Grunauer 1981, 287-288. 14 Säflund 1970, Paterake 2005, Becatti 1971. 15 Ashmole – Yalouris 1967. 16 Simon 1968, Stewart 1983, Herrmann 1987, Kyrieleis 1997, 2011. 13 © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 45-54 50 ANDRÁS PATAY-HORVÁTH debate. 17 It should be noted, that during the rearrangement of the fragments in the new archaeological museum of Olympia, the Greek specialists opted for another arrangement (Closed type “B”) and this solution was also advocated by the first detailed monograph dedicated to the reconstruction and interpretation of the east pediment. 18 The most recent volume on the other hand voted for Closed type “A”. 19 This one was suggested earlier by F. Studniczka, another archaeologist who used the life-sized plaster models at Dresden, and it was also advocated by P. Grunauer, an architect who has studied the temple for several decades and corrected the measurements (albeit by a few centimeters only) given by Treu for the dimensions of the pediment. 20 By doing this, he laid the foundation for every future reconstruction of the composition. In addition, a few years later, he made another important contribution to the reconstruction and following a detailed analysis, based exclusively on objective, measurable criteria, he concluded that from the four possibilities the reconstruction type “Closed A” is the least problematic. 21 structed statues. The digital models were produced by scanning the original fragments and by reconstructing them (i.e. completing their missing limbs and armour) virtually. 22 4.3 The virtual 3D reconstruction Since experimentation with the precious and monumental original fragments is out of question for practical reasons and plaster casts and models are expensive to produce and not easy to handle, it seemed to be reasonable to apply the latest 3D scanning technology to the problem. The aim of the project was to test the practical feasibility issues raised by the early experiments and to assess the aesthetic effects of the possible arrangements with 3D models of the reconBoth N. Yalouris (Ashmole-Yalouris 1967) and G. Becatti (Becatti 1970) have opted for the closed arrangement type A. Hurwit 1987: 6 note 2, Steuben 1990: 383, Knoll 1994: 80 just emphasize that the question is open to debate. 18 Säflund 1970, 81-96 with Fig. 56. 19 Paterake 2005, 171-174. 20 Grunauer 1971. 21 Grunauer 1981, 281-301, esp. 288. 17 Figure 8 Virtual 3D reconstructions of the central figures arranged as in Figure 7. The fragments are displayed in grey, the reconstructed parts in pale blue 22 For technical details see e.g. Patay-Horváth 2011a,b; 2012; 2013a. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 45-54 THE TEMPLE OF ZEUS AT OLYMPIA IN VIRTUAL REALITY 4.4 Results Studying and testing the different reconstructions with 3D models revealed that contrary to the expectations based on the results of the early experiments with plaster casts, every arrangement could be realized. (Figure 8) The virtual models show, however, that the arrangement, which was considered to be physically impossible in the 19th century (open “A”) and which is most commonly accepted today, is indeed the most difficult to realize (Figure 9): the limbs of figure K and G do not necessarily run across each other, but the distance between them is so small (max. 10 cm) that we can hardly believe that this arrangement could follow the original intentions of the designers or the sculptors. Furthermore, the model clearly shows that in the case of both open arrangements, another problem arises: the spears in the hands of the male figures fit the available space only if both of them grip the shaft directly under the spear-head (Figure 10) which is otherwise not attested in Greek art. In the case of closed arrangements (Figure11), we have no such problem with the spears; these arrangements can therefore be regarded more probable than the open ones. Though the remaining two closed arrangements are possible both technically and iconographically, one can observe, that every piece of evidence, which is independent from the interpretation actually points to type “A”, which can be considered therefore as the most probable reconstruction. 23 (Figure 12) 51 ent scholarly and educational purposes and it would be highly desirable to complete the digitization and virtual reconstruction of the temple’s architectural decoration (i.e. the west pediment and the 12 metopes depicting the labours of Heracles). Figure 9 Figures K and G according to the open arrangement Type “A” Figure 10 The spear-heads of the male figures in the open arrangements. 5. CONCLUSION The accurate virtual 3D reconstruction of the temple, including its east pediment and its cult statue as shown in Figure 13, has clearly demonstrated the possibilities and advantages of virtual archaeology. The digital models can be employed for differFor a detailed discussion see Patay-Horváth 2008. 23 Figure 11 The spear-heads of the male figures in the closed arrangement © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 45-54 52 ANDRÁS PATAY-HORVÁTH Figure 12 The virtual 3D reconstruction of the entire east pediment according to closed arrangement Type "A" Figure 13 The virtual 3D reconstruction of the entire east pediment according to closed arrangement Type "A" ACKNOWLEDGEMENTS The project was carried out with the financial support provided by the Norway Grants and the Hungarian National Research Fund (OTKA ref. no. NNF 85614) as well as the János Bolyai scholarship offered by the Hungarian Academy of Sciences. Special thanks are due to G. Hatzi (head of the Ephorate at Olympia) and R. Senff (German Archaeological Institute at Athens, supervisor of the Olympia excavations). I am indebted to Prof. B. Frischer (Univ. of Virginia) for his constant help and encouragement. Scanning of the original fragments of the east pediment was carried out by Tondo SP1 Ltd. (Budapest), the virtual reconstruction and the illustrations were produced by G. Gedei (occasionally assisted by D. Bajnok and M. Hitter). © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 45-54 THE TEMPLE OF ZEUS AT OLYMPIA IN VIRTUAL REALITY 53 REFERENCES Ashmole, B. – Yalouris, N. (1967) Olympia. The Sculptures of the Temple of Zeus. London: Phaidon. Becatti, G. (1971) Controversie olimpiche. Studi miscellanei Vol. 18, 65-84. Bulle, H. (1939) Der Ostgiebel des Zeustempels zu Olympia, Jahrbuch des Deutschen Archäologischen Instituts Vol. 54, 137–218. Dinsmoor, W. B. (1950) The Architecture of Ancient Greece: an Account of its Historic Development. 3rd ed. New York: Batsford. Gruben, G. (2001) Griechische Tempel und Heiligtümer, München: Hirmer. Grunauer, P. (1971) Der Zeustempel in Olympia – Neue Aspekte. Bonner Jahrbücher Vol. 171, 114–131. Grunauer P. (1981) Zur Ostansicht des Zeustempels. A. Mallwitz (ed.), X. Bericht über die Ausgrabungen in Olympia. Frühjahr 1966 bis Dezember 1976, Berlin: De Gruyter, 256–301. Heilmeyer, W.-D. et al. eds. (2012) Mythos Olympia. Kult und Spiele. Berlin: MartinGropius-Bau. Hellmann, M. C. (2002) L’architecture Grecque I. Paris: Picard. Hennemeyer, A. (2010) Der Zeustempel von Olympia, Griechenland. Geschichte der Rekonstruktion, Konstruktion der Geschichte, ed. by W. Nerdinger, München:Prestel, 218-219. Hennemeyer, A. (2011) Zur Lichtwirkung am Zeustempel von Olympia. P. I. Schneider, U. Wulf-Rheidt (eds.), Licht. Konzepte in der vormodernen Architektur (Diskussionen zur archäologischen Bauforschung 10), Regensburg: Schnell, 101–110. Hennemeyer, A. (2012) Der Zeus-tempel, Mythos Olympia. Kult und Spiele. ed. by W.-D. Heilmeyer et al., Berlin: Martin-Gropius-Bau, 121-125. Herrmann, H. V. (1987) Die Olympia-Skulpturen. Darmstadt: Wissenschaftliche Buchgesellschaft. Hurwit, J. M. (1987) Narrative Resonance in the East Pediment of the Temple of Zeus at Olympia, The Art Bulletin Vol. 69, 6-15. Kyrieleis, H. (1997) Zeus and Pelops in the East Pediment of the Temple of Zeus at Olympia. D. Buitron-Oliver (ed.), The Interpretation of Architectural Sculpture in Greece and Rome, Washington: National Gallery of Art, 12–27. Kyrieleis, H. (2011) Olympia. Archäologie eines Heiligtums. Mainz: Philipp von Zabern. Knoll, K. ed. (1994) Das Albertinum vor 100 Jahren - Die Skulpturensammlung Georg Treus. Dresden: Staatliche Kunstsammlungen. Lawrence, A. W. with revisions by R. A. Tomlinson (1983) Greek Architecture, New Haven: Yale Univ. Press. Lippolis, E. et al (2007) Architettura greca: storia e monumenti del mondo della polis dalle origini al V secolo, Milano: Mondadori. Mallwitz, A. (1972) Olympia und seine Bauten, Darmstadt: Wissenschaftliche Buchgesellschaft. Ohnesorg, A. (1993) Inselionische Marmordächer, Berlin: De Gruyter. Patay-Horváth, A. (2008) Zur Rekonstruktion und Interpretation des Ostgiebels des Zeustempels von Olympia. Mitteilungen des Deutschen Archäologischen Instituts, Athenische Abteilung Vol. 122, 161-206. Patay-Horváth, A. (2011a) Virtual 3D Reconstruction of the East Pediment of the Temple of Zeus at Olympia – A Preliminary Report”. Proceedings of the 14th International Congress “Cultural Heritage and New Technologies” 2009, edited by K. F. Ausserer, W. Börner, S. Uhlirz, Wien: Museen der Stadt Wien – Stadtarchäologie, 653-658. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 45-54 54 ANDRÁS PATAY-HORVÁTH Patay-Horváth, A. (2011b) The complete Virtual 3D Reconstruction of the East Pediment of the Temple of Zeus at Olympia“ Paper presented at 4th ARC 3D conference, Trento, Italy, 3-5 March 2011. Accessed 30th May 2012. http://www.isprs.org/proceedings/XXXVIII/5-W16/pdf/patay.pdf. Patay-Horváth, A. (2011c) The Virtual 3D Reconstruction of the East Pediment of the Temple of Zeus at Olympia (CD ROM), Budapest. ISBN 978-963-284-196-0 Patay-Horváth, A. (2012) Reconstructions of the East Pediment of the Temple of Zeus at Olympia – A comparison of drawings, plaster casts and digital models. International Journal of Heritage in the Digital Era, Vol. 1 No. 3, 331-349. Patay-Horváth, A. (2013a) Virtual 3D Reconstruction of the East Pediment of the Temple of Zeus at Olympia - An old puzzle of classical archaeology illuminated by recent technologies, Digital Applications in Archaeology and Cultural Heritage, http://dx.doi.org/10.1016/j.daach.2013.06.001. Patay-Horváth, A. (2013b) Die Perserbeute von Plataia, die Anfänge der elischen Münzprägung und die finanziellen Grundlagen der „Grossbaustelle Olympia”. Klio, Vol. 95, 61-83. Paterake, K. (2005) To anatoliko enaetio tou naou tou Dia sten Olympia. Rethymno: Panepistimo Kretes. Säflund, M. L. (1970) The East Pediment of the Temple of Zeus at Olympia. A Reconstruction and Interpretation of its Composition. Göteborg: P. Åström Schiering, W. (1991) Die Werkstatt des Pheidias in Olympia II, Werkstattfunde. Berlin: De Gruyter Simon, E. (1968) Zu den Giebeln des Zeustempels von Olympia. Mitteilungen des Deutschen Archäologischen Instituts, Athenische Abteilung Vol. 83, 147-167. Steuben, H.von (1990) Sterope und Hippodameia, Études et Travaux Vol. 15, 379-384. Stewart, A. F. (1983) Pindaric Diké and the Temple of Zeus at Olympia. Classical Antiquity Vol. 2, 133–144. Strocka V. M. (2004) Phidias, Künstlerlexikon der Antike, ed. by R. Vollkommer, München/Leipzig: Saur, Vol. II, 210-236. Studniczka, F. (1923) Die Ostgiebelgruppe vom Zeustempel in Olympia, Abhandlungen der Sächsischen Akademie der Wissenschaft, Phil.-Hist. Klasse Vol. 37, 3-36. Treu, G. (1897) Olympia III. Bildwerke aus Stein und Thon. Berlin: A. Asher. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 45-54 Mediterranean Archaeology and Archaeometry, Vol. 14, No 4, pp. 55-64 Copyright © 2014 MAA Printed in Greece. All rights reserved. ABOUT QUALITY AND PROPERTIES OF DIGITAL ARTIFACTS Călin Neamţu1, Daniela Popescu1, Răzvan Mateescu2, Liliana Suciu3, Dan Hurgoiu1 Department of Design Engineering and Robotics, Technical University of Cluj-Napoca, Muncii Avenue no. 103-105, Cluj-Napoca, 400641, Cluj, Romania 2 National History Museum of Transylvania from Cluj-Napoca, St. C. Daicoviciu, no. 2, 400020, Cluj, Romania 3 Department of Ancient History and Archaeology, Babeș-Bolyai University, St. Mihail Kogălniceanu, no.1, 400084 Cluj-Napoca, Cluj, Romania 1 Received: 01/12/2013 Accepted: 19/07/2014 Corresponding author: Călin Neamțu (calin.neamtu@muri.utcluj.ro) ABSTRACT Any form of use of virtual reality or augmented reality in history and archaeology is based on 3D digitized models that are obtained in various ways (3D modeling, 3D scanning, photogrammetry, etc.). These represent virtual replicas of real artifacts/monuments. In the authors’ vision, a digital artifact is represented by the digitized form of a historical artifact/monument. A virtual artifact is a concept that embodies not only the digital form, but also includes metadata, interactive elements, feedback elements, multimedia files, etc. coupled with stereo vision and the ability to interact with them by specific methods involving VR/AR. The quality of the 3D models used in AR/VR applications is influencing the visual experience of the users, and this represents a property of a virtual artifact that can be defined and quantified. The authors propose the introduction of the maximum permissible deviation term as the unit of measurement for the fidelity of the digitized 3D model. The quality of a 3D model does not depend only on the precision of the instrument/equipment used in the primary digitization phase and subsequent operations but also on operations prior to this step, such as mesh creation, surface creation, solid generation, optimization etc. The quality of the virtual/digital model is influenced not only by the methods used in order to obtain the 3D model, but also by the purpose for which it will be used (level of details are influenced by the limited amount of storage capabilities on some devices – AR). Other properties of a 3D model will be defined and exemplified, such as the traceability of the digital/virtual artifact, compatibility, interactivity and portability. The case study presented in this paper concerns the study of Dacian civilization from the Orastiei Mountains (the ancestors of the Romanian people) and represents the effort of an interdisciplinary team’s work. KEYWORDS: virtual artifacts, digital artifacts, virtual reality, quality CĂLIN NEAMȚU et al 56 1. INTRODUCTION The terms ”digital” and ”virtual” are attributes that nowadays are attached to an increasing number of entities in order to indicate their presence in a different than normal (physical) form – presented with the help of computer based technologies and multimedia, “characterized by electronic and especially computerized technology” [Oxforddictionaries, 2013] and stored in the form of numbers 0 and 1. As presented in "Tangible culture" [Hecher, 2012], cultural heritage institutions (galleries, museums and libraries) use more and more often digital media to present artifacts to their audience and enable them to immerse themselves in a cultural virtual environment [Bonis, B 2009] . A series of initiatives exist, both at a European as well as at a world-wide level, aimed at taking inventory and at digitally preserving the cultural heritage in either classical electronic form or in 3D format. Almost every project individually develops its own metadata structure to accompany a virtual/digital artifact but the initiatives that highlight the process and the method in which an artifact should be digitized in 3D format are most often independent and dispersed, as shown in [Cararae, 2013; Ronzino, 2012; Tien-Yu H, 2012; Popovici, 2008]. In this paper we present a series of properties which can be attached to a 3D digital/virtual artifact to record the creation history and traceability of a digital artifact that can be used both in research (digital 3D artifacts) and for the promotion of cultural heritage (virtual 3D artifact). Digitizing an artifact/monument can be done using several methods [Pavlidis, G, 2010, (Neamtu, et al. 2011,(a)]. The results vary from case to case, but the 3D model’s quality should be quantifiable in more ways than just through the number of polygons. The quality of a 3D model does not depend only on the precision of the instrument/equipment used in the primary digitization phase [Jakubiec, 2010] and subse- quent operations but also the post processing operation such as mesh creation, surface creation, solid generation, optimization etc., which can affect the quality of a 3D model. Using digital models for digital preservation of cultural heritage should be made to ensure that future generations have access to culture and history [Agosti, M., 2013], but these models are classified so that the correspondence between the physical and the virtual artifact can be emphasized through the metadata or quantifiable parameters. 2. DIGITAL AND VIRTUAL ARTIFACT In the authors vision the digital 3D artifact is a digital replica of a real artifact resulting from a rigorous digitization process through the utilization of scientific methods and instruments. A 3D digital artifact needs to record a history of all the operations needed for its modeling and to permit the intervention on this data at any moment. A 3D digital artifact is obtained through digitization that, most of the times, consists of 3D scanning of the real artifact, but other methods can also be employed such as photogrammetry for digitizing an artifact or monument. The virtual 3D artifact is a 3D model that demonstrates properties specific to virtual reality, that permits interaction by utilizing specific equipment and that can be stereoscopically visualized in an electronically generated environment or in a physical, real-world environment. A virtual artifact contains attached to the 3D model a series of interactive elements that permit the interaction between it, a virtual or real environment and the user. This, together with the visualization possibilities, makes the difference between a 3D digital artifact and a virtual one [Neamţu, et al. 2012 (b)]. A virtual artifact should be created starting from a 3D digital artifact that needs to be improved with specific virtual and augmented reality elements. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 55-64 ABOUT QUALITY AND PROPERTIES OF DIGITAL ARTIFACTS 3. QUALITY OF DIGITAL AND VIRTUAL ARTEFACT Two wide-spread definitions of quality state that quality is: “proper for use” (Juran, 2010) and “conformity with requirements” (Crosby 1995). If we take the above definition of quality and we try to quantify the quality of a digitized artefact or a virtual artifact we must define what mean “proper for use” and “conformity with requirements”. The artifact in digital format can be used in stand-alone applications (VR), VR web based or augmented reality applications. All of them are particular requirement for electronic form of real artifact. Finding a set of descriptors and indicators which can quantify the quality of electronic form of real artifact and store this descriptors / indicators in a set of metadata is a challenge and can be resolved only in a interdisciplinary approach. The same artifact should be able to be viewed in similar conditions on computers or smart devices using a variety of operating systems (cross-platform). What does it mean that "conformity with requirements" in case of a digital artifact? What are the requirements and what differentiates an artifact incorrectly digitized from a correctly digitized one? How can digital artifacts be classified based on elements that can remove the subjectivity of the classifier? These questions and problems can be solved only by identifying a number of parameters of the digitized model that can be uniquely determined and allowing the assignment of a numerical value to the quality of the digitization process. Digitizing a 3D monument / artifact can be done using several methods but laser scanning is probably the most accurate and wide-spread method that can be used today. Obviously, other methods such as modeling or photogrammetry cannot be ignored, but each may introduce a number of errors that are generated by the methods themselves which are known in the literature. In figure 1 and figure 2 are presented two different types of artifacts: a small one 57 (ceramic vessel) and a bigger one (a landscape with an ancient sanctuary). In case of the ceramic pot presented in figure 1 (a) the dimension of the artifact are 250mmx170mmx170mm and this artifact was digitized using laser scan (figure 1 b) a Original artifact b 3D scanned artifact d c 3D modeled artifact Comparison between scanned and modeled artifact Figure 1 Digitization of a ceramic vessel Figure 2 Large landscape digitization (sanctuary of Sarmisegetusa Regia) The landscape was digitized using photogrammetry. For understand why the precision is important in figure 1 (c), the 3D model of the ceramic pottery was exagger- © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 55-64 CĂLIN NEAMȚU et al 58 ated deformed and the deviation is presented in figure 1 (d). For this vessel a deviation of 10 mm is visible but for the landscape, 10 mm deviation may be the best result that can be obtained if we take into consideration its dimensions: 80mx20m. We propose the following set of parameters for quantify the quality of a digital artifact which describes the following properties: • accuracy and uncertainty of digitization • traceability • compatibility and portability • texture accuracy a) Real artifact b) Cloud points c) 3D Mesh d) 3D Surface 3.1 Accuracy and uncertainty of digitization In order to measure the accuracy and uncertainty of the digitization process (degree of fidelity), the authors propose the introduction of the term maximum permissible deviation (MPD): a value that highlights the maximum accepted deviation of the digital artifact compared to the real object. This can be done by comparison between the 3D model and the real artifact through CMM measurement or through scanning. To determine the deviation of the digitized model from the actual object the steps below must be followed: • The scans of the actual artifact; • The point cloud (Figure 3b) is converted into 3D mesh (Figure 3c) and then in surface (Figure 3d) using a CAD software; • The CAD model is imported into the CMM’s (Coordinate Measuring Machine) software for measurement (Figure 4); • Using the CMM’s software the real and the digital models are aligned with the help of established methods (PLP (point – line - plane), best fit, etc.); • Using the CMM’s touch probe and its software points of interest on the real artifact’s surface are recorded; Figure 3 Ceramic pottery digitization steps Figure 4 Comparison of real model with CAD model: top – procedure of points acquisition using CMM, bottom – graphical display of deviation © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 55-64 ABOUT QUALITY AND PROPERTIES OF DIGITAL ARTIFACTS • Using the CMM’s software the standard deviation is calculated between the real and the digital artifact (Figure 5). 59 can be optimized in order to improve them with new technologies and algorithms for generating 3D models. Figure 6 Deviation Analysis for two CAD models: point cloud and surface Figure 5 Graphic representation of point’s deviation using Histogram After comparing the CAD model to the real one, the value for the standard deviation is obtained. Next using a CAD tool (Deviation Analysis) we can check deviations recorded in each stage of the process of obtaining the CAD model. Thus the comparison of the two CAD models (e.g. the point cloud and mesh / surface) results in a deviation color map as presented in figure 6. Using the same instrument (Deviation Analysis) the differences between 3D models obtained by various methods of digitization can be determined, as shown in the figure 7. 3.2 Traceability The authors define traceability of the 3D digital artifact as its property to keep track of all operations that led to its creation and also of initial elements in their unaltered state - from point cloud to final model generation. If a digitized artifact keeps a documented history of operations used to generate it, any operation can be analyzed later and Figure 7 Deviation Analysis for two CAD models: point cloud and surface In general CAD software allows operation history preservation [Neamţu, et al. 2012], in the form of an organized tree, in Parent - Child structures (figure 8), in some situations, the operations can be sorted according to type, and not only in the order of their occurrence. Figure 8 shows the reconstruction of a ceramic vessel, in which the digitized fragments are clearly highlighted in both the textured and CAD model. For each scanned fragment distinct groups of wireframe elements (symmetry axis, boundary), surfaces or solids are created. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 55-64 CĂLIN NEAMȚU et al 60 Universal 3D formats (*.igs, *.stp, *.stl, *.u3d, etc.) are not capable of preserving the history of all necessary operations for 3D model creation, but they ensure the compatibility between different software platforms. But the 3D digital artifact model should be made available in these formats, in order for it to be used in applications, other than the ones in which it was built. Figure 9 Using AHP method for establishing the degree of importance for descriptors of traceability Table 1. Descriptors for traceability Element Figure 8 Reconstruction of a Dacian storage vessel (chiup) To assess the traceability of a digital artifact the authors propose using a set of descriptors associated with this property. The ranking of these descriptors was establishing using AHP method (Analytically Hierarchy Process) – figure 9. In table 1 is presented the ranking of descriptors resulted from AHP, classified in five categories: cloud points, mesh, wireframe, surface and solid. 3.3 Compatibility and portability Compatibility and portability of a 3D digital artifact can be defined as the properties that enable the model to be viewed and modified at all times, in a cross - platform way. Descriptor Points Record operation for generation of solid body Solid Record operations used for editing solid body Record assembly of solid body Record operation for generation of surfaces Record operations used for editing Surface surfaces Record operation for surface combinations (e.g. Split, Join, etc.) Record operation for wireframe generation Record operations used for editing Wireframe wireframe Record operation used for combine wireframe elements Record operation for mesh generation Record operations used for mesh Mesh editing Record operation used for combine mesh Record operation for cloud points filtering Cloud Record operation for cloud points Points editing Preservation of initial cloud points 4.9 4.5 2.2 5 5 2.9 2.5 2.3 2 11.3 13.3 9 12.6 11.9 15.6 Cross – platform means using different operating systems and hardware configurations for both computers (Figure 9 top) and for smart devices (Figure 9 bottom). Thus, a 3D digital artifact should ensure the possibility to be viewed (and edited if it is possible), by using different software from the one it was created in, even if the history of all necessary operations for its creation cannot be preserved (when transferring the 3D model, using standard CAD formats). © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 55-64 ABOUT QUALITY AND PROPERTIES OF DIGITAL ARTIFACTS 61 ing a 3D model can be done by mapping or when scanning using a color laser during the digitization process. Faithful representation of the texture can provide important information not only about color, pigment or painting technique but also in terms of the method with which the vessel was created, the quality of the material or the burn process. Figure 11 Difference between real and digital textured model (top) and UV map texture for pottery (bottom) Figure 10 Distribution of operating system (OS) for computers (top) and for smart devices Using standardized formats for 3D models it is possible that the digitized artifact can be viewed on various hardware and software configurations. Another problem to be solved is how to view a model with older or newer versions (of software) than the one that created the initial file. For example the files with the extension *.IGES (Digital Representation for Communication of Product Definition Data), which began to be standardized in 1980 (U.S. National Bureau of Standards - NBSIR 80-1978), can be used for 3D data manipulation on most common OS and CAD software and maybe is the most used 3D format for sharing 3D models between CAD platforms. To assess the compatibility and portability the authors propose a quantitative assessment based on the number of CAD formats in which the 3D model can be exported, directly from the application in which it was generated. 3.4 Texture accuracy Texture is the element that gives the visual appearance of digital artifacts. Textur- Mapping a simple photograph onto the 3D model may not lead to the best result (Figure 11). Using a more advanced method like UV texture mapping (Cube mapping, MIP maps, etc.) which associates each pixel to each physical area of the artifact can lead to better results. Another method for automatic association between the 3D model and texture is laser scanning of the artifacts. This method usually creates automatic mapping using UV or UVW methods. Also photogrammetry techniques can yield satisfactory results as shown in Figure 12. Figure 12 Melancholic Roman (Delphi Archaeological Museum): digital model obtained using photogrammetry (left) and real artifact (right) © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 55-64 CĂLIN NEAMȚU et al 62 Evaluation for the texture in terms of quality is quite difficult because, its quality is directly influenced by the method of acquisition, lighting conditions, or equipment used in the process. For an objective evaluation of this, the authors believe that a minimum number of parameters must be known so that subsequent processing of color can be made on the basis of measured parameters during data acquisition. Table 2. Descriptors for texture accuracy Parameter Equipment type Software version Lens Exposure Time Resolution Luminous intensity ISO Value Cannon D550 1.0.9 18-55mm 1/30sec 18Mpx 18600 lux 160 Because in both cases, for laser scanners that can acquire texture and other equipment, the texture is acquired through an optical sensor. The parameters describing the acquisition of a texture should reflect its quality and the environmental conditions in which the data acquisition was made as shown in Table 2. 4. CONCLUSION AND FUTURE WORK In this paper the authors propose a set of metadata that help define and establish the technical quality of digital artifacts. The notion of quality of a historical or cultural artifact in the approach of this paper refers only to the quality of digitization and digital processing of the artifact and does not consider their historical or artistic qualities. Documenting the technical quality of a historical artifact is very important in terms of long-term preservation, so that in the long-term the digitization of a historical artifact ensures the preservation of accurate information about it even if the actual artifact suffers degradation or irreversible damage. Figure 13 Points filtering and mesh generation, versus CAD Deviation Analysis in case of Dacian iron artifact Parameters chosen by the authors to describe the technical quality of a digital artifact are descriptors that can be calculated mathematically (standard deviation where accuracy and uncertainty of digitization) or can be physically determined (traceability or texture accuracy). The term of maximum permissible limit of deviation (MPD) is inserted: a value that highlights the maximum accepted deviation of the digital artifact compared to the real one and two methods are presented which permit the determination using measurement with CMM or with Deviation Analysis with a CAD instrument (Figure 13). From the point of view of long-term preservation techniques we can attempt to determine some parameters that describe the technical quality of a digitized artifact. The number of points and triangles can provide an overview regarding the fidelity of the digitized surface (Figure 12) but how can the size of those triangles affect the quality of the artifact and the file size space required. The authors identified a number of issues which still need to be researched and for which they are still seeking solutions. One of the research directions is trying to answer the questions: How to measure texture mapping? How to measure the quality of a texture in comparison with the real texture of an artifact? © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 55-64 ABOUT QUALITY AND PROPERTIES OF DIGITAL ARTIFACTS The authors try to establish a procedure or a formula in the form of: Q= (Reqp + μ + Stdev + …..) / (….) where: • Reqp - resolution of equipment, • μ - uncertainty of digitization process, • Stdev - standard deviation of the final model from scan result, • ….. – other parameters. This approach attempts to determine a mathematical formula to comprise more parameters and to provide a single score for the technical quality of an artifact. From the point of view of the traceability of a digital artifact, the recording of all the necessary operations needed for its creation will allow in the future the reinterpretation of this data using new algorithms to generate the 3D models. Ensuring the compatibility and portability should be a requirement for digital artifacts because modern technology (hardware and software) is very diversified and has a rapid rate of development. The availability in digital format of an artifact for researchers and general public is an objective which must be bore in mind when digitization is fulfill. As shown in [Forte, M. 2011], there is a strong gap between data capturing and data accessibility, which in our opinion can be decreased if we define compatibility and portability as property of digital artifact. This two 63 property can be evaluated and measured and can be an indicator for accessibility to the digital and virtual form of an artifact. The texture of a digital artifact is probably as important as the 3D shape when it transmits artistic or historical information. The digital preservation and the correct overlay over on the 3D model are very important considering the time and the conditions of exposure or storage of historical artifacts which invariably affects the texture over time. Attaching some information regarding the acquisition manner of the texture in digital format and properties of the sensor used may allow subsequent preparation of the digital texture in a way in which the *.raw format (called also digital negative) allows intervention on light intensity and color of the scene. The measurement of digitization precision of the texture can be a very laborious and expensive process that at this time is practically impossible to do for each digitized artifact. Therefore, the authors propose attaching a data set that allows recreating the conditions in which the texture was digitized in the idea that in the future these conditions could be recreated in the virtual environment and the texture will be corrected if necessary. ACKNOWLEDGEMENT This paper was supported by the project "Progress and development through postdoctoral research and innovation in engineering and applied sciences – PRiDE - Contract no. POSDRU/89/1.5/S/57083", project cofounded from European Social Fund through Sectorial Operational Program Human Resources 2007-2013. REFERENCES Agosti, M. (2013) Experiences and perspectives in management for digital preservation of cultural heritage resources (Panel). Communications in Computer and Information Science Vol. 354, 2013, 1-3, ISBN 978-3-642-35833-3 Bonis, B., J. Stamos, S. Vosinakis, I. Andreou and T. Panayiotopoulos (2009) A platform for virtual museums with personalized content. Multimedia Tools and Applications Volume: 42 Issue:(2), 139-159 ISSN: 1380-7501. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 55-64 64 CĂLIN NEAMȚU et al CARARE (2013) Metadata schema: http://www.carare.eu/eng/Resources/CARAREDocumentation/CARARE-metadata-schema, accessed October Crosby, P., (1995) Philip Crosby's Reflections on Quality. McGraw-Hill. ISBN 0-07-014525-3. Forte, M. (2011). Cyber-Archaeology: Notes on the simulation of the past. Virtual Archaeology Review, 2(4), 7-18 Hecher, M., Möstl, R., Eggeling, E., Derler, C., Fellner, D.W., (2012) Tangible culture Designing virtual exhibitions on multi-touch devices ELPUB 2012 - 16th International Conference on Electronic Publishing Jakubiec, W. and W. Płowucha (2010) Methodology for uncertainty estimation of coordinate measurements, 10th International Symposium on Measurement and Quality Control 2010, ISMQC 2010, Osaka, ISBN 9781617820199 Joseph M Juran (2010) Juran's quality handbook, New York, ISBN ISBN:9780071629737 McGraw Hill Neamtu, C., D. Popescu and R. Mateescu (2011) From classical to 3D archaeology. Annales d'Universite 'Valahia' Targoviste, Section d'Archeologie et d'Histoire Vol.13, Issue 1, 79-88 ISSN: 15841855. Neamtu, C., R. Comes, R. Mateescu, R. Ghinea and F. Daniel (2012) Using virtual reality to teach history." Proceedings of the 7th International Conference on Virtual Learning, 2-3 Noiembrie, Brasov, Romania, 303-310, ISSN 1844-8933. Neamţu, C., Z. Buna, R. Mateescu, F. Popişter and X. Morar (2012) CAD software in 3D reconstructions of ancient world. Quality - Access to Success Vol. 13 Issue:(SUPPL.5), 513-518 Oxforddictionaries – definition of digital, available on line at http://oxforddictionaries.com/definition/american_english/digital Pavlidis, G., Koutsoudis, A., Arnaoutoglou, F., Tsioukas, V., Chamaz, C. (2007) Methods for 3D digitization of Cultural Heritage, Journal of Cultural Heritage, Volume 8, Issue 1, pp.93-98; Popovici, D.M. Crenguta. M. B., Matei, A., V. Voinea, N. Popovici, (2008) Virtual Heritage Reconstruction Based on an Ontological Description of the Artifacts. International Journal of Computers Communications & Control Vol. 3, 460-464, ISSN: 1841-9836. Ronzino, P., Hermon, S., Niccolucci, F., (2012) A Metadata Schema for Cultural Heritage Documentation, in V., CApellini (ed.), Electronic Imaging & the Visual Arts: EVA 2012 Florence, Firenze University Press, 36- 41. StatCounter - GlobalStats (2013) Top 7 Operating Systems on August 2013 – available online at http://gs.statcounter.com/#os-ww-monthly-201308-201308-bar StatCounter - GlobalStats (2013) Top 8 Mobile Operating Systems on August 2013 – available on-line at http://gs.statcounter.com/#mobile_os-ww-monthly-201308-201308bar Tien-Yu, H. (2012) A unified content and service management model for digital museums International Journal of Humanities and Arts Computing. Vol. 6, 87-99 DOI 10.3366/ijhac.2012.0040, ISSN 1753-8548. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 55-64 Mediterranean Archaeology and Archaeometry, Vol. 14, No 4, pp. 65-74 Copyright © 2014 MAA Printed in Greece. All rights reserved. ARCHFIELD: A DIGITAL APPLICATION FOR REALTIME ACQUISITION AND DISSEMINATION – FROM THE FIELD TO THE VIRTUAL MUSEUM Neil Smith1 and Thomas Levy2 1Visual Computing Center, King Abdullah University of Science and Technology (KAUST), Saudi Arabia 2Department of Anthropology and Qualcomm Institute, University of California, San Diego, USA Received: 30/11/2013 Accepted: 17/07/2014 Corresponding author: Neil Smith (neil.smith@kaust.edu.sa) ABSTRACT The lack of efficient digital data processing tools during field excavations is a major bottleneck affecting the delay between data collection and dissemination in archaeology. In this paper, we outline the fundamental methodology of ArchField, an integrated digital field recording solution developed to overcome this bottleneck and translate field excavations to virtual museums in real-time. ArchField records sub-centimetre accurate three-dimensional coordinates from Total Stations and RTK GPS units. Recorded field data and measured 3D coordinates are digitally processed to produce auto-generated daily GIS top plans. The processing pipeline enables the generation of publishable online maps from the first day of excavation to the last. It is interoperable with many different GIS viewers and stores data in an online PostGIS database. Digitization of archaeological data in the field is streamlined to facilitate standardization, redundancy and storage that can be immediately made accessible online to the digital community. Consequently, ArchField integrates features such as synchronization, data formatting, re-projection, dynamic labeling and symbolization. It provides immediate online accessibility of field excavations for virtual museums of the future. ArchField enables any archaeological project to inexpensively adopt real-time 3D digital recording techniques in their field methods. KEYWORDS: GIS, real-time, southern Jordan, LiDAR, SfM. 66 SMITH & LEVY 1. INTRODUCTION As the humanities moves towards a digital domain of data sharing and analysis, the long delay between field excavation and public dissemination becomes an increasing problem. A primary bottleneck is insufficient data processing, vetting, and database tools to manage the vast amounts of data as it is recovered on a daily basis from field excavation. Instead of reducing the workload of excavations, digitization without the proper data handling tools in place can easily become almost intractable in size and complexity (see Petrovic et al. 2011). The end result is a significant portion of time spent during post excavation to process data into a workable form that only then can be adequately analyzed and visualized. Errors made during digital field recording may not be caught until much later and at this point cannot always be easily resolved. Since archaeologists have only one chance to excavate an area and the only data that can be analayzed are what was properly recorded, it is essential that development of efficient and comprehensive digital recording and processing tools are developed. In this paper, we present ArchField as a computational solution for efficiently recording, analysing, and modelling the various sources of data recovered on a daily basis from field excavations. Fundamental to the methodology behind ArchField is a focus on automated digital processing to streamline and speed up the archaeological recording process while providing automatic and user-informed data vetting tools while still in the field. Archfield provides a unified software to combine high precision spatial recordings (Survey and LiDAR/SfM) with supervisor’s observations and digital spreadsheets. Integrated databases are seamlessly synced between the field excavations and lab analyses to enable raw data from the field to be immediately visualized as 3D top plans and queryable field reports in real-time. It is a digital application that serves as a bridge between field excavations and spatially oriented visualization and virtual presentation. ArchField enables any archaeological project to inexpensively adopt real-time 3D digital recording techniques in their field methods. It is field tested having undergone three excavation seasons of development and beta testing in southern Jordan (Smith and Levy 2012). We present its current developments and its interoperability with different archaeological recording schemas, excavation methodologies, measurement instruments and other 3D digital acquisition tools such as LiDAR and Structure-from-Motion (SfM). Figure 1 ArchField running on the iPad. 2. RELATED WORK Several joint computer science and archaeological projects have sought in the past to develop software to digitally process data after excavations have occurred. For example, DATARCH (Fabricatore and Cantone 2007) developed in 2006 functioned as an image management system where different media, primarily digital photos, could be stored and connected to a relational database. ArchaeoloGIS (Montesinos et al 2010) and ETANA-GIS (Gortan et al. 2006) were designed as open-source GIS map servers that could take basemaps (generated in ArcGIS© from traditional excavation techniques) and database tables (recorded in Access© or other spreadsheet programs) and serve them on online. 3DMurale is another application that stores data in SQL tables, plans to develop tools for on the field recording, and provide a full visualization package including the use of SfM (Green et al. 2002; Van Gool et al. © University of the Aegean, 2014, Mediterranean Archaelogy & Archaeometry, 14, 4 (2014) 65-74 67 ARCHFIELD: A DIGITAL APPLICATION 2002; http://dea.brunel.ac.uk/project/murale/). Unfortunately funding ended for the project in 2003 and only the database for Sagalassos, Greece was implemented. REVEAL an NSF funded computer vision project (Gay et al. 2010) for archaeology is a recording tool that combines plan reports with continuous video recording of excavations and multi-view camera captures of important artifacts and features. The future goal of the project is to orient surfaces and artifacts in 3D space using techniques such as multi-view stereo. Similarly many projects have adopted off-the-shelf proprietary GIS software (e.g. ArcGIS) or open source GIS programs (e.g. MapInfo, GRASS, OSSIM, QGIS) to digitize their data and publish studies in scholarly journals or in online databases (e.g. Harrower 2010; Al-Kheder et al. 2009; Ross et al. 2005). Others have used not only surveying tools but also LiDAR or SfM (e.g. Al-Kheder et al. 2009; Allen et al. 2004; Forte 2013; Pollefeys et al. 2003) to document excavations. However, what all these projects lack is software to facilitate integration of their data entry with survey instruments while still in the field. The development of software to record detailed provenience and descriptor data in the field that directly communicates with high precision recording equipment has remained a project of the commercial sector. Two of the more well-known proprietary data entry software that archaeologists have used are Solofield developed by TDS and ArcPad developed by ESRI. Both of these programs have been developed for surveying and thus are not as easily adaptable to archaeological recording. The major drawback for archaeologists is that there is still little data processing occurring during excavations. The recorded results must still be downloaded, imported into GIS software and then manually processed to create top plans and final maps. They fall short of providing to the field of archaeology a fully implemented solution for high precision data recording, data organization, visualization, and analysis. 3. METHODOLOGY ArchField builds upon a decade of previous methods of digital field recording (Levy and Smith 2007) and has now undergone several major revisions over the past three years as it has been thoroughly field tested in Southern Jordan. Its evaluation has led us to address in the software’s recording methods four main requirements of digital field excavation: 1) Precise survey instruments must seamlessly integrate with digital excavation data entry in the same application; 2) The software should intelligently facilitate the recording and storage of data with high redundancy, reduction of user intervention, standardization, and remote accessibility; 3) Various vetting tools, automated labelling, and real-time visualization of the on-going excavations must be provided to archaeologists while in the field; and 4) Diverse datasets such as 3D scanning and aerial mapping must be automatically integrated for visualization and digital dissemination. 3.1 Integration of High Precision Survey Tools with Archaeological Data Entry The primary functionality of ArchField is to facilitate and streamline the procedures to properly record the provenience of artifacts and loci in 3D space and directly link them with their detailed field descriptions. Data entry for artifacts consists of 3D recording (x, y, and z) their unique locations using survey tools and simultaneously storing associated metadata (i.e. basket identifier, provenience, date, classification, description, etc.). 1 In contrast to most excavations that still rely upon imprecise survey recording methods such as tape measures and periodic dumpy level 2 elevation readings to For a detailed description of all the excavation methods applied please see Levy and Smith 2007. 2 Elevation readings can be precise up to 2.0mm when a digital auto level is used. However, not all excavations take a reading for every artifact but assign the elevation of the locus. The 1 © University of the Aegean, 2014, Mediterranean Archaelogy & Archaeometry, 14, 4 (2014) 65-74 68 SMITH & LEVY plot artifact or loci onto graph paper, ArchField uses a Leica total station as the primary surveying tool. It takes precise 2.0mm + 2 ppm readings of artifacts and loci that are immediately stored and displayed on a digital Top Plan. When the user presses record on the Total Station, the measurements are directly read into ArchField. The raw (x,y,z) distance measurement is combined with the known position of the total station and a spatial reference system to project the 3D measurement into a coordinate that can be located on a map or in GIS software. The spatial reference system used to store the coordinates is Universal Transverse Mercator (UTM). Within the ArchField software it is possible to convert on-the-fly the coordinate into any other spatial reference system such as WGS84 Lat/Long as used by Google Earth. When the coordinate is received from the total station, the supervisors’ entered information is combined and stored together in ArchField’s database. For artifacts, the most critical data entry is the provenience information that describes to which basket, locus, square, and area an artifact belongs. In our excavations the basket number (e.g. 5000…5001…5002) is a sequential number assigned to each artifact. The basket number becomes a unique identifier otherwise known as a ‘key’ in the relational database that can be used to retrieve all the tabular data associated with that artifact, including its coordinate location (see 3.2 for a description of the DBMS architecture and schema). Loci are digitally recorded as threedimensional polygons with the ability at every change in depth to create a new representative polygon and updates to the metadata. A locus is defined here as a distinguishable layer of soil deposit in which artifacts and other features are found. Loci are demarcated in excavations by the volumetric space of their dumpy level provides only precision on the zaxis and does not account for the imprecise reading on the x,y coordinate of a measured artifact using measuring tape. depositional layer. A locus in three dimensions can be represented as a polyhedron, but typically it is drawn on graph paper as a boundary of the layer’s extent. In order to digitally record a locus, we use the total station to take multiple position readings along the physical boundary of the locus. The readings from the total station are automatically connected together in ArchField as vertices of a three-dimensional polygon. As its excavated depth increase we can generate a polyhedron to represent its threedimensional nature (fig. 2). Each locus is assigned a unique number similar to the basket number assigned to artifacts. The extensive tabular data collected on a locus (e.g. sediment composition, density, types of artifacts, associated features, stratigraphic relationships, excavation strategies) are all linked to this unique locus number. Figure 2 Loci extruded as color coded polyhedrons representing opening and closing elevations (visualized in ArtifactVis2). The supervisor is assisted in taking detailed information on each artifact or locus using ArchField’s data entry form. It allows the supervisor to enter all the pertinent information of the artifact or locus and have it automatically combined with its 3D location using the integrated total station surveying tool. To streamline the data entry interface and reduce possibilities of mis-entered data, the application is designed to auto-complete as much information as possible. Once the first artifact/locus is recorded the data entry form remains populated with information that does not change from one artifact to the next. After each artifact is recorded the © University of the Aegean, 2014, Mediterranean Archaelogy & Archaeometry, 14, 4 (2014) 65-74 69 ARCHFIELD: A DIGITAL APPLICATION basket number is incremented to reflect the following unique basket number. Drop down buttons are automatically populated with supervisors’ agreed upon descriptions of artifacts or locus characteristics. The general descriptions of the artifacts remain standardized across multiple excavation areas and save the supervisor time from manually re-entering the same description for artifacts. The user only needs to change on a regular basis one or two fields saving time in the long run and preventing typical data entry mistakes. Whenever an artifact or locus is recorded a table appears below the data entry form showing all the pertinent features recorded. This serves as a final check that is used to confirm that all the information was entered correctly prior to moving on to the next recording and allows immediate edits to be carried out if mistakes are found. In this manner, ArchField streamlines the user’s entry of data and simplifies user-assisted correction. Currently, three different versions of ArchField have been designed to address the evolving landscape of emerging computer hardware and operating systems. In 2010, ArchField was designed as a web based version using the combination of HTML, PHP, and Javascript languages (fig. 3). Figure 3 ArchField web version Top Plan with OpenLayers integration. The main advantage of this approach is that it can be run on any operating system with a web browser and the code can be easily changed without a need to recompile. The web based version has the most minimal hardware requirements. It can be deployed on an Atom or ARM based pro- cessor with 1GB of ram and only consumes 100mb of file storage. Essentially, it can be run on any computing device that can serve a webpage. This means the web based version can run on a tablet, smart phone, imbedded device, netbook or standard laptop. Recently ArchField has been ported to run as a native iOS app (fig. 1) and is currently being rewritten to run as an osindependent C++ compiled version to enable more complex features on Windows based tablets (fig. 5). The advantage of ArchField as an os-independent GIS and data entry tool is that it can be deployed on any device rugged enough to be brought out to the field. 3.2 A Database Management System with high redundancy, standardization, and remote accessibility Archfield integrates PostGIS a SQL relational database management system. PostGIS was chosen due to its broad adoption by various GIS applications (e.g. QGIS, GRASS, ArcGIS) and the Open GIS Consortium (OGC), its rich feature set of GIS functions, and ability to fully serve web based mapping systems. PostGIS provides a robust and efficient query system allowing asynchronous spatial queries across a shared network. The DBMS architecture is designed to maintain a synchronized dataset across local and remote servers. As data is recorded in the field, it is stored on the field lab’s server computer and on a periodic basis synchronized with an online server hosted back at the university. 3 ArchField handles the synchronization by merging the local and remote DBMS using SQL protocols and checks prior to any merger for conflicts. Any updates made during synchronization are recorded in a separate table with a timestamp allowing a simplified With the field lab’s 3G internet connection, it is possible to synchronize all the databases in real-time, but in practice synchronization is conducted only on a periodic basis to save bandwidth. 3 © University of the Aegean, 2014, Mediterranean Archaelogy & Archaeometry, 14, 4 (2014) 65-74 70 SMITH & LEVY method of version control to be implemented. This method allows continuous harmonization of the data stored locally and remotely over time. ArchField’s database is a distributed system where multiple users are able to access and change entries in the DMBS without the creation of divergent copies. The DBMS schema organizes the spatial information and supervisor field entered data into two primary tables: one for artifacts and another for loci. Other tables generated from the analysis of the artifacts or loci are relationally joined to these primary tables (for a description of the complete DBMS schema see Gidding et al. 2013 and Smith et al. 2013). The tabular relationships of the DBMS allow ArchField to later perform SQL based queries across the entire DBMS and render the results as a 2D or 3D map. No matter whether in the field, in the dig lab, or back at the university the server database can provide supervisors and specialists real-time access to data being recorded in the excavations. ArchField provides a project tailored setup to handle diverse recording methodologies of different archaeological sites. During the creation of a new excavation project in ArchField, users are given the option to setup what field descriptors, unique identifiers, and database attributes (columns) are pertinent to their recording of artifacts and loci. ArchField takes this information to generate a tailored database and data entry interface for the project. 3.3 In the field Vetting, Tagging, and Visualization An essential aspect of the creation of ArchField is the ability to automatically process incoming data so that it can be viewed in real-time as digital top plans (figure 2). Dynamically changing data can be handled so that a Top plan’s symbology, labels and colors are auto-generated. ArchField is designed to constantly update auto-generated KML 4 files as the PostGIS database receives new entries, changes, and undergoes time (top plans must reflect the current day of excavation rather than all artefacts or loci exposed over the whole season). The symbology (symbols to distinguish different types of artifact), color schemes (unique colors to differentiate soil types) and labeling (key metadata to assist the supervisor) imbedded in the KML is created on the fly whenever a new point is created. The creation of dynamic KML files allows Archfield to be interoperable with most programs that support KML. These programs become real-time GIS spatial views of that day’s excavation as if several hours were spent preparing a top-plan after the day’s excavation had occurred. This is in contrast to a GIS program simply reading the spatial database, because imbedded in the KML file is not only the spatial and metadata but also specific instructions on how to render each individual artefact and locus. The latest version of ArchField imbeds OpenLayers, an Open Source pure JavaScript library for mapping data in twodimensions. Unlike Google Maps, OpenLayers does not require an internet connection to function but like Google Map, MSN Virtual Earth, Bing, etc. when there is internet connectivity it can stream all of the same imagery data. OpenLayers also enables full exploitation of the tablets’ and web based version’s multi-touch interfaces. The digital top plan is designed to allow the field supervisor and registrar to visualize and vet their excavations as they would on a traditional paper top plan but with the added benefit of a full GIS. They are able to conduct queries, toggle layers, make quick edits, and zoom in on pertinent features. They see the current day’s top plan as it is KML is an OSG approved vector format supported among other GIS programs (i.e. Google Earth, QGIS, GRASS, ArcGIS,etc). KML files created in ArchField can be opened in any of these other programs. 4 © University of the Aegean, 2014, Mediterranean Archaelogy & Archaeometry, 14, 4 (2014) 65-74 71 ARCHFIELD: A DIGITAL APPLICATION constructed and can immediately tell whether a locus or artifact was incorrectly recorded. The real-time top plans allow an archaeological project’s supervisor to catch mistakes while still in the field, being more in-tune with the current process of digital recording, and be freed up to focus on other aspects of documentation after a morning of excavation. Having real-time and vetted top plans by the end of the excavation enables immediate analysis and a direct transition to final publishable excavation plans. ArchField’s ability to automatically curate spatial data into a comprehensive GIS makes it the essential backbone to later virtual museums and 3D visualization. One other technique to reduce error and streamline field recording is ArchField’s ability to generate printed labels with barcodes for every basket (Figure 4). Once an artifact or basket elevation is recorded a button appears for label printing. When this button is clicked it pulls the information from the table and prints a label with all the important information on an encoded barcode. The printed labels save time for the registrar, eliminate the chance that they may write the wrong information, and prevent the label from being misread by others. During laboratory hours, the label and barcode further reduce possible error in the whole process. The barcode stores a unique identifier enabling that specific entry to be located in the database. After a day’s excavation, the lab supervisor uses their bar code reader to quickly ‘check in’ all the artifacts and buckets collected that day from the field. In this way, before the data may be used by other lab specialists it is triple checked to make sure there is no error in the data. Finally, when a find is scanned it is recorded as received from the field and where it is currently stored (e.g. washing, conservation lab, photography lab, 3D scanning lab, pottery lab, storage, etc). When the sorted artifacts are moved to storage in a crate, the label is scanned again to update the database. By this manner, a detailed final list of where all artifacts are stored can be printed at the end of the excavations. Figure 4 digitally generated label with barcode. 3.4 Integration of 3D Scanning, and Visualization The application of LiDAR and Structurefrom-Motion to archaeology has opened the avenue to easily capture 3D models of site architecture and stratigraphic levels. These techniques are integrated with ArchField to meet the goal of a total 3D documentation of the field excavations. The ability to scan a complete excavation in its full three-dimensions provides a context to visualize the three-dimensional artifact positions and locus boundaries recorded by total-stations and GPS. In our recording methodology, ArchField is used to generate the ground control points to geo-reference 3D scans (e.g. LiDAR and SfM) so that the resulting point cloud models can be loaded into the same geographic space as the recorded artifacts and loci. SfM is employed to supplement the LiDAR scans where there are occluded or inaccessible areas and provide more frequent capture of locus surfaces and changes as the site is excavated (Figure 5). The initial geo-referenced LIDAR point-cloud is used as the reference for all future SfM recordings so that as new scans are produced they can be registered in proper position. The latest version of ArchField has been rewritten in C++ and uses OpenSceneGraph to efficiently render these point cloud models onto the recorded Top plan. Although in its initial alpha stage, ArchField C++ is able to render our most dense SfM models and triangulated meshes on Windows Surface Tablets at full 60 frames per second (Figure 5). Every component of the data recorded are immediately availa- © University of the Aegean, 2014, Mediterranean Archaelogy & Archaeometry, 14, 4 (2014) 65-74 72 SMITH & LEVY ble for analysis and visualization both in ArchField and our 3D visualization software called ArtifactVis2 (see Figure 7; Smith et al. 2013). These procedures enable the accurate digital reconstruction of the site’s architecture and significant artifacts. The intention of the reconstruction is not necessarily to show artistically how the site may have looked at one point in time but provide a faithful three-dimensional record of the excavations for on-going analysis. This has allowed us to continually return to the site examining its architecture, spatial distribution of artifacts and stratigraphic layers in a fully immersive 3D environment that is connected to the same GIS server that ArchField updates on a daily basis in the field. We have been able to show other archaeologists the excavations and discuss in detail various aspects of the excavation process and theories on its use from across the globe. Figure 5 ArchField C++ with 3D point cloud of Khirbat al-Iraq excavations. 4. FIELD EVALUATIONS From 2010 through 2012, ArchField was evaluated by site supervisors and staff at five sites in Southern Jordan dating from the Early Bronze to Islamic periods. In this paper, we focus on the most recent excavations at Khirbat Faynan in Southern Jordan. Evaluations of the software discussed below are based upon the application of the software in the field excavations and the written comments of the field supervisors. In 2012, Khirbat Faynan one of the largest sites in lowland Edom with extensive occupation during the Iron Age, Roman, Byzantine and Islamic periods was exca- vated using ArchField. During this season, the iOS and improved web version were tested. In contrast to past seasons (20022008), where often the supervisors had to stay behind to complete recording artifacts at the site with the Total Station, ArchField allowed the supervisors to keep up with the excavation process and finish on time. Several evaluators specifically noted that the label printing in the field removed the tedious writing out of labels. In addition, top plans were complete at the end of the daily excavation, requiring ca. 10-15 minutes in the lab to synchronize the field database with the main lab server and print paper copies of the Top Plans. 5 A general conclusion found in all the evaluations was that ArchField enabled a new level of efficiency in survey tool integration, top plan generation and lab workflow. After testing the iOS version of ArchField the staff evaluations provided very useful suggestions of how to adapt and improve on the system in future versions. First, there was an agreement that the lighter weight of the iPod and the ability to use it in sunlight outweighed the benefits of the larger screen on the iPad. Smartphones and other handheld devices are preferable in field data entry. Second, although the purpose of the iPod version of ArchField was to enable the supervisors to become more mobile, the observed practice of the users was that they remained next to the total station and registrar table even though they were untethered. Finally, evaluations found that a remaining limitation of ArchField was still a need to communicate with the Total Station and label printer through a laptop since neither could be supported directly through the iPad. ArchField requires a somewhat complex initial setup and involves many components (batteries, transformers, and RS232-to-USB converts) to integrate the label printer and total station which when In the past (Levy and Smith 2007) an average of 1-3 hours was spent importing artifact and loci recordings into ArcGIS and manually generating the Top Plan. 5 © University of the Aegean, 2014, Mediterranean Archaelogy & Archaeometry, 14, 4 (2014) 65-74 73 ARCHFIELD: A DIGITAL APPLICATION not communicating properly introduced problems in the field-workflow. The main feature request by the supervisors was a completely wireless solution, with easy setup, and could communicate directly with surveying instruments and the label printer without the need of extra hardware, power sources, or cables. 4. CONCLUSIONS WORK AND computers, online and in virtual museums. ArchField is a computational solution to translate field excavations to virtual museums and catalogues in real-time. FUTURE The greatest outcome from using ArchField in the last several excavations is how it has allowed us to disseminate our data through scientific visualization with minimal postprocessing after excavations. Every component of the data stored in ArchField has been immediately available for analysis and visualization in ArtifactVis2 (see Figure 6; Smith et al. 2013) and ArchaeoStor (Gidding et al. 2013). This has allowed us to continually return to the site examining its architecture, spatial distribution of artifacts and stratigraphic layers. It has enabled us in our own research to digitally present our ongoing research to a large audience of archaeologists on our individual Figure 6 ArchField artifacts and loci displayed in same geographic space as SfM and LiDAR scans (visualized in ArtifactVis2) Over the three seasons of ArchField’s use, mobile and scanning technology has significantly changed. Future research will be directed towards reducing components, going fully wireless and integrating more extensively SfM into ArchField with daily recording of surfaces, loci, and in situ artifacts. ACKNOWLEDGEMENT This project was funded in part by a NEH Digital Humanities Start-up grant, NSF IGERT, and King Abdullah University of Science and Technology. We are grateful to the Department of Antiquities of Jordan, Dr. Mohammad Najjar, and the ELRAP/L2HE staff and students. REFERENCES P. Allen, et al. (2004) Digitally modeling, visualizing and preserving archaeological sites. Proc. Joint Conference on Digital Libraries 2004 (JCDL 2004). Tuscon, AZ. June 7–11, 2004, 2004. Fabricatore, G. and F. Cantone. (2007) Pushing the Archaeological Interpretation by Analysing Workflow Protocols: The “Variable Transparency Image Stacker” and DATARCH© Archaeological Data Management System . Layers of Perception – CAA 2007. Forte, M. (2013) Virtual Reality, Cyberarchaeology, Teleimmersive Archaeology. In Campana S., Remondino F., 3D surveying and modeling in archaeology and cultural heritage. Theory and best practices. BAR ARCHAEOPRESS, Winter, 2013. Gay, E., K. Galor, D. B. Cooper, A. Willis, B. B. Kimia, S. Karumuri, G. Taubin, W. Doutre, D. Sanders, and S. Liu (2010) REVEAL Intermediate Report. Proceedings of CVPR Workshop on Applications of Computer Vision in Archaeology (ACVA'10), June, 2010. © University of the Aegean, 2014, Mediterranean Archaelogy & Archaeometry, 14, 4 (2014) 65-74 74 SMITH & LEVY Gidding A., Y. Matsui, T. E. Levy, T. DeFanti, and F. Kuester. (2013) ArchaeoSTOR: A Data Curation System for Research on the Archaeological Frontier. Future Generation Computer Systems 29:2117 – 2127. Gorton D., R. Shen, N. Srinivas Vemuri, W. Fan, E. A. Fox. (2006) ETANA-GIS: GIS for Archaeological Digital Libraries. In JCDL'06, June 11–15, 2006, Chapel Hill, North Carolina, USA. Green D., J. Cosmas, T. Itagaki, M. Waelkens, R. Degeest, E. Grabczewski. (2002) A real time 3d stratigraphic visual simulation system for archaeological analysis and hypothesis testing. In Virtual Archaeology Proceedings of the VAST2000 Euroconference held in Arezzo, November 2000.Oxford, Archaeopress. Harrower, M. J. (2010) Geographic Information Systems (GIS) hydrological modeling in archaeology: an example from the origins of irrigation in Southwest Arabia (Yemen). Journal of Archaeological Science 37:1447-1452. Al-Kheder, S., Y. Al-shawabkeh, and N. Haala. (2009) Developing a documentation system for desert palaces in Jordan using 3D laser scanning and digital photogrammetry. Journal of Archaeological Science 36:537-546 Levy, T. and N.G. Smith. (2007) On-site digital archaeology: GIS-based excavation recording in southern Jordan. In Crossing Jordan--North American Contributions to the Archaeology of Jordan, T. E. Levy, M. Daviau, R. Younker and M. M. Shaer, Eds. London: Equinox, pp. 47-58. Montesinos, J.F, C.L. Lopez, F. M. Rangel Pardo. (2010) ArchaeoloGIS: Using Geographic Information Systems to Support Archaeological Research. In COM.Geo 2010, June 21-23, 2010 Washington, DC, USA. Petrovic, V. A., A. Gidding, T. Wypych, F. Kuester, T. A. DeFanti, and T. E. Levy. Dealing with archaeology's data avalanche. IEEE Computer Society, July 2011, pp. 56-60. Pollefeys, M., L. Van Gool, M. Vergauwen, K. Cornelis, F. verbiest, J. Tops. (2003) Imagebased 3D Recording for Archaeological Field Work. Computer Graphics and Applications (CGA). 23(3): 20-27. Ross, K.A, A. Janevski, and J. Stoyanovich. (2005) A Faceted Query Engine Applied to Archaeology. In Proceedings of the 31st VLDB Conference, Trondheim, Norway, 2005. Smith NG, and Levy TE. (2012) Real-time 3D archaeological field recording: ArchField, an open-source GIS system pioneered in Jordan. Antiquity 85(331):on-line http://antiquity.ac.uk/projgall/smith331/. Smith, N. S., K. Knabb, C. DeFanti, P. Weber, J. Schulze, A. Prudhomme, F. Kuester, T. Levy, and T. DeFanti. (2013) ArtifactVis2: Managing real-time archaeological data in immersive 3D environments. Proceedings of IEEE 19th Int’l Conference on Virtual Systems and Multimedia (VSMM 2013). L. Van Gool, M. Pollefeys, M. Proesmans, A. Zalesny. (2002) The MURALE project: image-based 3D modeling for archaeology. In Virtual Archaeology Proceedings of the VAST2000 Euroconference held in Arezzo, November 2000.Oxford, Archaeopress. © University of the Aegean, 2014, Mediterranean Archaelogy & Archaeometry, 14, 4 (2014) 65-74 Mediterranean Archaeology and Archaeometry, Vol. 14, No 4, pp. 75-81 Copyright © 2014 MAA Printed in Greece. All rights reserved. MUSEUMS AND SOCIAL MEDIA: MODERN METHODS OF REACHING A WIDER AUDIENCE Panteleimon Marakos 15th Ephorate of Byzantine Antiquities, 11 Socratus Str., 69100, Komotini, Greece Received: 30/11/2013 Accepted: 22/06/2014 Corresponding author: pantelis_marakos@yahoo.gr ABSTRACT The opening of museums to society brought about radical changes in the museum practice, because their goal is not only the viewing of exhibits but a meaningful contact and communication with the public. According to this view, museums are trying to approach a wider audience, providing them with the opportunity of personalized use of information and active creation of content in an entertaining and interactive way. The following study explores various approaches that the social media (Facebook, Twitter, Flickr, Linkedin) can provide to museums, aiming at a constant communication and interaction with the audience. The vertiginous technological development, the digitization, the dissemination and democratization of knowledge, as well as the systematic information of the public by the mass media, have significantly influenced the museums in the way they promote their activities. The social media can be low-cost communication tools while addressing to a wider audience, as they can provide museums with the opportunity to benefit in many ways from their use, offering them the ability to give prominence to both their dynamic nature and the purpose of their actions. By studying the cases of important museums in Australia, America and Europe, it becomes immediately clear that the social media have already formed a basic communication tool for the museum’s exhibitions and for the other traits that highlight the educational and entertaining dimension of their character. KEYWORDS: museums, social media, communication, participation, audience. 76 PANTELEIMON MARAKOS 1. INTRODUCTION The opening of museums to society brought about radical changes in the museum practice, because their goal is not only the viewing of exhibits but a meaningful contact and communication with the public. According to this view, museums are trying to approach a wider audience, providing them with the opportunity of personalized use of information and active creation of content in an entertaining and interactive way. The following presentation explores various case studies of how museums have used social media, aiming at a constant communication and interaction with the audience. Today, social media are considered to be the heart of the internet, as they foster and enhance communication, participation, diffusion of information and feedback among users which provides a powerful means of sharing, organizing and finding content and contacts. In addition, social media are organized around users, whereas the Web is largely organized around content (Mislove et al. 2007, 1). Among the most popular social media are Facebook, Twitter, MySpace, YouTube and Flickr. In order to have an account on social media, someone has to create a profile on the platform of their choice, by providing some basic information about themselves. The majority of people who use social media are considered to be youngsters between the age of 15-35. According to the degree of their participation, we can divide the users into three categories (Kumar et al. 2006, 616): • passive users, are those who join the network out of curiosity, but never engage in any significant activity, • practical users/inviters, are those who created a profile for a specific reason and recruit people to participate, • active users, who are full participants in the growth of the online social network, and actively connect themselves to other members. The rapid development of social media and the continuous increase of the number of users have led companies, corporations, organizations and institutions to exploit the applications and the possibilities they provide, with the aim to advertise and promote their products and services (Evans & McKee 2010, 4). Museums are also starting to use social media in order to reach a wider audience, by building an interactive environment that will foster the public to engage in discussion, through sharing information and content. Despite the fact that there isn’t any significant amount of writings and research for museums that use social media, there are plenty of case studies that enlighten us about the utility and the role they play in the communication policy of a museum. 2. AUSTRALIAN MUSEUM The first case study indicates the importance of social media in creating the context of a periodical exhibition, through the constant participation of the audience. The “All About Evil” exhibition concept came to the Australian Museum from the Royal Tropical Institute (Tropenmuseum) of Amsterdam after the success of their exhibition displayed in 2006. The exhibition was built from the Tropenmuseum’s cultural collections and included over 900 items, having loans from European collections and private lenders (Kelly 2009, 9). While an interesting topic, it was a controversial one with provocative connotations and potentially graphic subject matter. With this in mind, the Australian Museum did some preliminary work with the audience to gauge reactions to the overall topic, as well as feedback about some of the material that may be displayed (Jensen & Kelly 2009, 20-21). In order to conduct some further evaluation it was decided to use social media to both engage potential audience and compare this approach to a more ‘traditional’ front-end evaluation process. In February 2009 an “All About Evil” Facebook group was created, in order to test whether Face- © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 75-81 MUSEUMS AND SOCIAL MEDIA book would provide a vehicle for discussion on themes and possible content for the audience (http://www.facebook.com/home.php?ref =home#/group.php?gid=63750884739). Participants embraced the tools of Facebook, even contributing photographs and tagging photos uploaded by Museum staff. The group proved to be popular, gaining over 200 members in the first three weeks and generating a great deal of activity and discussion between the Museum and members, as well as among members themselves (Jensen & Kelly 2009, 22). According to Lynda Kelly, Head of Web and Audience Research of Australian Museum at that time, the use of social media is proved to be an effective and inexpensive way for the museum to participate in an open dialogue with the audience (Kelly 2009, 9). 3. CURRIER MUSEUM OF ART A similar case study comes from the Currier Museum of Art in Manchester, where it was developed a creative social media and email marketing program to engage with members and visitors, to encourage museum attendance and also to help shape an entire exhibition. In 2010, the Currier’s exhibition, The Secret Life of Art: Mysteries of the Museum Revealed, provided visitors with a behind-the-scenes look at how the staff cared for, displayed, and stored its 31,000 works of art when not in use. Developed as a “living dialogue” the exhibition was meant to answer the questions museum staff had received from visitors on a regular basis (Stern 2011, The Currier Museum of Art 2011). Social media was a critical piece of promoting this exhibition and keeping visitors engaged. During the exhibition’s opening, the museum asked members and guests to tweet their thoughts about both the opening and the exhibition itself, using the Twitter. Encouraging the opening’s attendees to tweet, not only helped build buzz around the exhibition when it debuted in October 2010, it also helped generate 77 greater awareness of the Currier’s own Twitter handle. In addition, YouTube videos, a dedicated blog, and related articles in the Currier’s monthly email newsletter, were also developed to round out the physical exhibition (Stern 2011). In short, the results were astounding. Over the four months the exhibition was open to the public, the Currier Museum grew its Twitter following by 49%, while expanding the number of ‘Likes’ on its Facebook Page by 24%. Plus, the museum added more than 700 names to its email list (Stern 2011). 4. POWERHOUSE MUSEUM The next case study, points out how social media can extend the authenticity and credibility of a museum by enabling it to maintain a cultural dialogue with its audience in real time. Part of the Powerhouse Museum of Australia, the Sydney Observatory responded to a then-current Web rumor that planet Mars would be unusually close to the Earth. The senior curator posted this comment (Russo et al. 2007, 22): There is an email circulating in cyberspace saying that the red planet Mars will be exceptionally close on 27 August (2006). According to one version “It will look like the Earth has two moons”!!! Once again this is a good lesson in not believing everything on the Internet. The email is a hoax…(Lomb 2006). Over the next month, one hundred and thirty five visitors of the blog responded to this comment. Some examples of their comments include (Russo et al. 2008, 24): Ah, I thought the email was a little too exaggerated to be true...Thanks to the Observatory for setting the record straight and informing the public (Eve Aug 19th, 2006 at 6:01 pm). Ah ha …. it sounded too good to be true and I headed straight on over to the “professionals” here at the Sydney Observatory to set my mind at ease that the email is as STUPID as I thought it sounded!... Thanks Sydney Observatory…. (Koobakoop Jul 27th, 2006 at 1:26 pm). Thanks for explaining this so clearly. My six-year-old is still awake at 9:20 p.m. waiting up to see Mars between 10 p.m. and 3 a.m. Tom came home from school on Friday excited about © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 75-81 78 PANTELEIMON MARAKOS the coming event. I thought it sounded too good to be true, punched it into Google 10 minutes ago, showed him your site, and he’s on his way to bed!! [David, Aug. 27, 2006]. It is not insignificant that many of the responses to the Senior Curator’s comments credited the Sydney Observatory with providing the “truth” in this matter. This example illustrates how social media can be used to enable a cultural and scholarly dialogue while strengthening the veracity of museum knowledge. The subsequent communication demonstrates how the many-to-many model can enhance both audience interaction, experience and museum authority (Russo et al. 2007, 22). 5. SMITHSONIAN INSTITUTION Another case study explores how the Smithsonian Institution is using social media to enlist the public in delving into its collections, expanding its research and, sometimes, just adding interesting postscripts to history. Crowd-sourcing and user-generated content are not new to the Smithsonian. In 1850s, Joseph Henry, one of the Smithsonian’s first secretary, enlisted volunteers around the country to gather observations about storms and other weather occurrences and to telegraph them to Washington. This initiative helped the Institution to establish weather patterns and maps, which later evolve into the National Weather Service (Olson 2011, Bray et al. 2011). These days, the institution is embracing social media to involve the public by helping the Smithsonian solve puzzles like identifying some early 20th-century women. In March 2009, the Archives posted eight unidentified photographs from the Science Service Collection on Flickr. Tammy Peters, Supervisory Archivist, wrote a blog post, asking for visitors to help identify these women. By July, 2009, Flickr user, "rockcreek," had identified Elizabeth Sabin Goodwin by citing a link to a newspaper clipping (Olson 2011, Bray et al. 2011). In April 2010, Elizabeth Goodwin's granddaughter, Linda Goodwin, recog- nized the Flickr photo and contacted the Archives through Flickr, contributing details of her grandmother’s life and some of her drawings that allowed Smithsonian Archives to advance its research on the Science Service collection. As blog commenter Penny Richards, summarized, "To see this story go from an image with initials to a full biography with images and living memories, through crowd-sourcing, is wonderful, one of the very coolest parts of the whole Flickr Commons project for me" (Olson 2011, Bray et al. 2011). The results were (Olson 2011, Bray et al. 2011): • Four of the eight unknown women in the Science Service Collection were identified with verifiable sources. • The Smithsonian Archive's audience was increased • The Flickr Commons has successfully increased traffic and resources to the Science Service Finding Aid. 6. BROOKLYN MUSEUM The next example shows how the Brooklyn Museum began to evaluate various Web 2.0 sites to see how they could help them create more interactive exhibition content. The Interpretive Materials plan for Graffiti invited museum-goers to ‘tag’ two designated walls within the exhibition space. This wall was the tipping point for their collaboration, going forward, and the catalyst for building a dynamic interactive Web site (Caruth & Bernstein 2007). Anticipating that this wall would change significantly during the exhibition’s eightweek run, the Interpretive Materials manager planned to take the activity one step further - inviting visitors to track the progress of the “Museum Mural” on their Web site. At the suggestion of members of the Information Systems department, they employed the popular photo-sharing site Flickr, and for $24 a year (the cost of a Flickr pro account), they could upload new digital images, weekly, taken off the wall. Through the use of Flickr, they realized that they could provide this activity quick- © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 75-81 MUSEUMS AND SOCIAL MEDIA ly and efficiently, and without the arduous in house development of their own program (Caruth & Bernstein 2007). Now that they had their Flickr account, they started to think about other ways it could be leveraged. They decided to create another project in which they asked visitors to submit images of already-existing street art from their local Brooklyn neighborhoods. Photos e-mailed to the Museum were posted to their Flickr account throughout the run of the exhibition. In the end, the community established the firstever Museum archive of local street artists (Caruth & Bernstein 2007). In addition to promoting the Graffiti exhibition on their own Web site, they got the word out through their MySpace page. Interestingly, one of the local street artists, Ellis G. (Gallagher), who was recorded for their podcast series, had a popular MySpace presence. With Ellis’s help on MySpace, the Museum was able to gain good word-of-mouth about the exhibition and publicize the gallery talks. Along the way, Ellis helped the Museum make many friends in MySpace (Caruth & Bernstein 2007). The plan was a success in terms of sheer numbers, and an even bigger success in terms of mission (Caruth & Bernstein 2007): • The Graffiti exhibition in Brooklyn Museum was viewed 12,376 times • 913 photographs of graffiti were submitted on Flickr • 1,338 virtual drawings were submitted in Brooklyn archive • The Museum’s MySpace page had over 3,000 friends. Moreover, they had discovered that community on the Web didn’t necessarily mean programming on their own site. On the contrary, seeking out audience in their own Web communities (Flickr, MySpace) was even more powerful. 79 July 2011 two PhD students, Matthew Law (Cardiff University) and Lorna Richardson (University College London), began an online social media experiment with the title Day of Archaeology (Pett 2011). The Day of Archaeology project aimed to provide a window into the daily lives of archaeologists around the world. The project asks people working, studying or volunteering in the archaeological world to participate, by recording their working day and sharing it through text, images or video via the use of social media. All these submissions were moderated and released through the project’s website and disseminated through different social media networks, such as Flickr, Facebook, and Twitter (Pett 2011). The project had expressions of interest from people working on excavations, scientists working in laboratories, archaeologists talking about how cuts have affected their work, community archaeologists leading workshops and museum educators teaching the next generation about the magic of archaeology. When it was completed, the experiment formed part of Lorna’s PhD research and also was written up for academic publication and was used as a model for public engagement at the 2011 Theoretical Archaeological Group conference in Birmingham (Pett 2011). The resulting Day of Archaeology project demonstrates the wide variety of work their profession undertakes day-to-day across the globe, and helps to raise public awareness of the relevance and importance of archaeology to the modern world through the use of social media. The first ever Day of Archaeology was held on July 2011 and had over 400 contributing archaeologists and due to the success, the project continued in 2012 and also in 2013 (Day of Archaeology 2012). 8. CONCLUSIONS 7. DAY OF ARCHAEOLOGY The last but not the least case study underlines the contributory role of the individuals through the use of social media. On To sum up, the use of social media offer greater scope for collaboration, enabling museums to respond to changing demographics and psychographic characteris- © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 75-81 80 PANTELEIMON MARAKOS tics of the public. Recent years have seen more and more museums embracing social media, with particularly the Englishspeaking countries taking the lead. The examples aforementioned represent a shift in how museums interact with the public, whereas social media could be used to • develop new models of participation and feedback • promote museum’s activities • extend the authenticity and credibility of a museum • shape an exhibition’s content. However, many questions remain for researchers, designers, and practitioners, including (Russo et al. 2007, 22): • How much does the museum invest in revealing knowledge held in the community? • How far willing is the museum to relax its own authority in these areas of knowledge? • To what extent is the museum willing to promote community knowledge over its own? • Wouldn't it be better to target specific audiences? An additional challenge for museums is to consider the relationship between online and physical visitors, and what characteristics and behaviors may be shared across both. In order to get a deeper understanding of social media usage among museums, further studies and research are needed. Social media is not just about opening up another marketing channel, but it enables audience’s participation in many levels. The study has shown that museums consider and use social media to a high degree, as a means to attract more visitors to onsite museums. Despite the fact that online museum communities for both visitors and professionals are not yet very well developed in general, the examples in this paper illustrate a number of important first steps, whereas personalization will be an important aspect of most such efforts. REFERENCES Bray, P. Dalton, J. Dietrich, D. Kapsalis, E. Springer, M. and Zinkham, H. (2011) Rethinking Evaluation Metrics in Light of Flickr Commons. In Museums and the Web 2011: Proceedings, J. Trant, and D. Bearman (eds.), http://www.museumsandtheweb.com/mw2011/sessions.html, (15/02/2014). Caruth, N. and Bernstein, S. (2007) Building an On-line Community at the Brooklyn Museum: A Timeline, In Museums and the Web 2007: Proceedings, J. Trant, and D. Bearman (eds.), http://www.museumsandtheweb.com/mw2007/papers/caruth/caruth.html, (17/02/2014). Day of Archaeology, (2012) A day in the life of archaeologists: About the project, http://www.dayofarchaeology.com/about-the-project/, (14/02/2014). Evans, D. and McKee, J. (2010) Social Media Marketing. The next generation of business engagement, Indianapolis, Wiley Publishing. Jensen, B. and Kelly, L. (2009) Exploring Social Media for Front-End Evaluation. Exhibitionist, vol. 28, No 2, 19-25, http://name-aam.org/resources/exhibitionist/backissues-and-online-archive, (14/02/2014). Kelly, L. (2009) The Impact of Social Media on Museum Practice. In Proceedings of the Social Media and Museum Education Workshop, 1-14, http://australianmuseum.net.au/document/The-Impact-of-Social-Media-onMuseum-Practice, (18/02/2014). © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 75-81 MUSEUMS AND SOCIAL MEDIA 81 Kumar, R. Novak, J. and Tomkins, A. (2006) Structure and Evolution of Online Social Networks. In Proceedings of the 12th ACM SIGKDD international conference on Knowledge discovery and data mining, 611-617, New York, ACM. Mislove, A. Marcon, M. Gummadi, P.K. Druschel, P. and Bhattacharjee, B. (2007) Measurement and analysis of online social networks. In Proceedings of the 7th ACM SIGCOMM conference on Internet measurement, 29-42, New York, ACM. Olson, E. (2011) Smithsonian Uses Social Media to Expand Its Mission, http://www.nytimes.com/2011/03/17/arts/design/smithsonian-expands-itsreach-through-social-media-and-the-public.html, (15/02/2014). Pett, D. (2011) A day in the life of a lot of archaeologists, http://blog.britishmuseum.org/tag/social-media/, (16/02/2014). Russo, A. Watkins, J. Kelly, L. Chan. S. (2007) Social media and cultural interactive experiences in museums. Nordic Museology, vol. 1, 19-29, http://www.nordiskmuseologi.org/English/ANGELINA%20RUSSO.pdf, (17/02/2014). Russo, A. Watkins, J. Kelly, L. and Chan. S. (2008) Participatory Communication with Social Media, Curator, vol. 51, No 1, 21-31, http://onlinelibrary.wiley.com/doi/10.1111/j.21516952.2008.tb00292.x/abstract , (17/02/2014). Stern, A. (2011) How the Currier Museum is using social media, http://www.socialbrite.org/2011/05/05/how-the-currier-museum-is-usingsocial-media/, (15/02/2014). The Currier Museum of Art, (2011) The Secret Life of Art: Mysteries of the Museum Revealed, http://www.currier.org/exhibitions/secret-life-art-mysteries-museumrevealed/, (14/02/2014). © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 75-81 Mediterranean Archaeology and Archaeometry, Vol. 14, No 4, pp. 83-91 Copyright © 2014 MAA Printed in Greece. All rights reserved. ENGINEERING CAD TOOLS IN DIGITAL ARCHAEOLOGY Zsolt Buna1, Daniela Popescu1, Radu Comes1, Ionuț Badiu1, Răzvan Mateescu2 Department of Design Engineering and Robotics, Technical University of Cluj-Napoca, Muncii Blvd, no. 103-105, 400641, Romania 2 National History Museum of Transylvania from Cluj-Napoca, C. Daicoviciu St, no. 2, 400020, Romania 1 Received: 30/11/2013 Accepted: 16/07/2014 Corresponding author: Zsolt Buna (zsolt.buna@gmail.com) ABSTRACT This paper presents an original approach in the virtual reconstruction of destroyed ancient monument types which have no similarities to other standing monuments, or documentation about the design of these constructions. In this case the virtual reconstruction is a challenging act, which can be done using a large variety of designs. These designs must be validated not just from an archaeological or historical, but also from an engineering point of view, to create valid virtual models from the construction’s point of view. For this reason the authors chose to create the virtual reconstructions in Computer Aided Design (CAD) environment, in the detriment of design software’s, because it’s easier to create the virtual models and to do the simulations in the same software environment. Also in this paper are presented different reconstruction methods (photogrammetry using profile drawings and knowledge database creation) that can be achieved with the use of CAD software. The authors used different levels of details (LOD), which can be helpful in the validation process, where can be observed the structure of the monuments very detailed, and for renderings for dissemination. The case study was conducted on a destroyed Dacian watch tower by an interdisciplinary team composed of archaeologists, historians and engineers. The reconstruction of the watch tower was carried out in CATIA V5, and was disseminated through a video render, an virtual reality website and was imported into an augmented reality application. KEYWORDS: 3D reconstruction, 3D model, ancient destroyed monument. 84 ZSOLT BUNA et al 1. INTRODUCTION Since technology progresses very fast and computers become more powerful, they are used in various fields of science. A new trend, digital archaeology, also uses these technologies of computer graphics, computer aided design and digital imaging. A number of three-dimensional reconstructions and models of monuments and landscapes have been created in the context of culture and archaeology. The quality of such models is always questionable in terms of their resolution and detailed information they contain, as well as the concepts, references and background information that they are based upon (Theodoropoulos, Moullou et al. 2009). In addition, cultural heritage can benefit from high accuracy three-dimensional digital imaging for conservation, study and restoration of works (Berndt and Carlos 2000). Apart from being easier to interpret than two-dimensional drawings, these models facilitate data necessary for reconstruction projects, preservation or rehabilitation of the architectural or archaeological heritage (Amparo Núñez, Felipe Buill et al. 2012). Within this context, any action of digitization and virtual reconstruction of monuments to enrich the digital cultural heritage is welcome. Some virtual restorations of ancient monuments have been already made with different software solutions (see Table 2). Table 2 Reconstructions and software solutions. Virtually restored monument Abbey of Pomposanear Ferrara, Italy Amra palace, Jordan Software used Polyworks Polyworks Bam citadel, Iran 3DS Max Roman villa of Casal de Freiria, Portugal AutoCAD Reference (El-Hakim, Beraldin et al. 2002) (Al-kheder, Al-shawabkeh et al. 2009) (Futragoon, Kitamoto et al. 2011) (Rua and Alvito 2011) In the case of the above mentioned reconstructions, the monuments were intact or just partially damaged, so the process of reconstruction is easier because the design of the buildings is well known. In the case where there is little information (ancient descriptions or hand drawings) available about the destroyed monuments and there are no intact similar monuments after which to create the reconstruction, the process of virtual rebuilding is difficult and can have various outcomes with many different designs. The authors chose to create the reconstructions in a user-based CAD environment, CATIA V5, because can offer assistance in several stages of the reconstruction like: creating standardized construction elements and finite element analysis to validate the structure of the monument. The presented methodology includes steps such as monument digitization, processing and optimization of threedimensional model, simulation and dissemination of the virtual reconstructed monument using virtual reality and augmented reality technologies. In this methodology the CAD software is used especially to validate the design of the virtually reconstructed monument, but also to create the three-dimensional model, because it’s easier to create and simulate a model in the same software environment. 2. HISTORICAL CONTEXT Poiana lui Mihu watch tower (near Orăştioara de Sus village, Hunedoara County, Romania) is part of the Dacian fortification system from Orăştie Mountains included on the UNESCO heritage list. From archaeological sources it appears that, in the mid first century BC, the Dacian people built the first fortification structures of limestone blocks, according to Hellenistic influence techniques. Archaeological research has shown that this is a quadrilateral tower, which has medium dimensions compared to the other towers within Orăştie Mountains. Its outer side measures 11.75 meters, its inner side © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 83-91 ENGINEERING CAD TOOLS IN DIGITAL ARCHAEOLOGY 6.15 meters and its wall thickness it’s about 2.8 meters. 85 architects, specialists in related fields, like chemistry, materials science, virtual reality. Since the archaeological data can be very difficult to understand and to interpret by non-specialists, the team setup Figure 1 Illustration of a Hellenistic type of wall. According to (Gelu 2012) the lower part of the tower consists of strings of limestone blocks (four rows in most of the segments, three rows of blocks on some portions of the wall), but the upper part, which consisted of a structure made of wood and very compacted clay, was affected by the passing of time. During the archaeological excavations were found several fragments of tiles, fallen especially outside of the tower’s walls. 3. METHODOLOGY For virtual reconstruction of destroyed historical monuments the authors propose the use of computer aided design software solutions instead of the computer graphics software solutions. CAD software solutions can provide support on several stages of the virtual reconstruction of a monument. In the case where there is little archaeological information or it is missing, different simulation modules or finite element analyses can be used to verify various hypotheses. The methodology set up by the authors for the development of valid virtual threedimensional reconstructions of monuments can be seen in Figure 5. In the authors methodology the virtual reconstruction of a destroyed historical monument should be done by interdisciplinary teams to ensure the credibility and the accuracy of the virtual reconstruction. Depending on the nature of the reconstructed items the team can consist of: historians and archaeologists, engineers and Figure 5 Virtual reconstruction algorithm using CAD software solutions for destroyed monuments. is the first and the most important phase of this methodology. In this phase are gathered the team members who have different areas of expertise, and whom will help in the gathering and interpretation of the information, which is needed for the virtual reconstruction. The next phase is the documentation, where all the information is gathered from archives and from the fields. This information consists of ancient descriptions about the monument, hand drawings, older and recent archaeological finds. All of this information is systematized for further use by the team. In the next phase the engineers or those who will create the three-dimensional models will check if they have enough information to complete the task of virtual reconstruction. If the virtual reconstruction is not possible with the current information the team has to gather more information or the team needs to be extended with specialists from related fields. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 83-91 86 ZSOLT BUNA et al In the next phase are created the virtual three-dimensional models using all the information collected in the previous phase. In this step the engineers, the archaeologists and the historians are working together to ensure the historical rigor of the virtually reconstructed models. Using CAD software several model creation techniques are available, like general threedimensional modeling, modeling using photogrammetry or with creating a knowledge database with parameterized construction elements. The creation of parameterized elements starts with the creation of a base model which contains all the possible features that an element can have (gauges, cut-outs, holes, etc.), which are parameterized for further use in the knowledge database. Parameterizing a model means that for the base model’s dimension or angles are assigned standard dimensions or angles (which are average usual dimensions or angles found by the archaeologists during excavations), which later can be modified using the user interface of the database by the user. Also for different features are assigned Boolean operations, in general “false”, in this way is generated the simplest model (because for false Boolean operator the element is not shown), but also the user can change all the features (selecting true for the Boolean operator) using the interface of the database to his or her likings. Another method is used for generating three-dimensional models, which is similar to photogrammetry techniques, using a detailed hand drawing on paper, done by the archaeologists on the dig site. This drawing is digitized using a paper scanner for further use in the CAD software. Using this method a fairly accurate reconstruction of the watch tower’s wall can be created, but only on two axis: X and Z, on the Y axis due to the wall’s construction the dimensions cannot be measured without destroying the monument, therefore are selected randomly, knowing the usual dimensions of these limestone blocks from previous archaeological finds. Archaeological wall drawings also help in the creation of a database, because the limestone blocks can be measured on the drawings, also different types of blocks can be identified and categorized. To ensure the scientific rigor and to validate on a scientific level the virtual reconstruction a series of engineering analyses will be conducted on the created threedimensional models. If the created three-dimensional models pass the analyses they are presented for other scientists from related fields (other archaeologists, historians, architects, engineers, etc.), who will validate the virtual reconstructions. In this phase it’s important to have validation from other scientists, because they can see the created reconstruction from another point of view and can raise new questions or problems, which were overlooked by the initial team. If the validation is a success, the archaeologists and the historians can create the metadata, in text, video, sound and picture formats, for the three-dimensional model, which will be used in the dissemination process. In order to import the generated threedimensional models into augmented reality or virtual reality applications the raw models have to be optimized to reduce their sizes and shape complexity. The authors propose the internet as the primary channel for dissemination, in virtual reality applications, because it is an accessible technology available for the general public. 4. CASE STUDY The methodology described above has been already validated by the authors in various applications regarding virtual reconstructions of destroyed ancient monuments belonging to the Dacian civilization from the Orăștie Mountains. The proposed methodology above has been used as follows. Team setup - In the case of the Dacian watch tower’s virtual reconstruction the interdisciplinary team consisted of experts from the National History Museum of Transylvania and the Technical University of Cluj-Napoca, Romania. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 83-91 ENGINEERING CAD TOOLS IN DIGITAL ARCHAEOLOGY Documentation – At this stage all the published books, articles, archaeological surveys, drawings, pictures regarding the Greek, Roman and Dacian watch towers were gathered by the archaeologists, curators and historians. Documentation analysis – In this stage the engineers analyse the information package created in the previous step. They decide if the virtual reconstruction of the Dacian watch tower is possible with the given information. 3D model creation - The authors used CATIA™ V5 engineering CAD software solution, developed by Dassault Systèmes, to create the detailed three-dimensional model of the Dacian watch tower. The authors used different levels of detail in the virtual reconstruction of the tower to show that different levels of detail can be achieved using engineering computer aided design software platforms. A portion of the tower’s wall was virtually reconstructed with the help of photogrammetry (see Figure 6). Figure 6 Virtual reconstruction of a portion of the tower’s wall using a profile drawing. Another portion of the tower’s wall (see Figure 7), the wooden wall and the roof structure was reconstructed using a knowledge database (see Figure 8), which contains parameterized construction elements, such as limestone blocks, connection beams, beams with special cut-outs and heat threated ceramic tiles. Using a database to construct the three-dimensional virtual monument speeds up the process and provides a very detailed reconstruction. 87 Figure 7 Virtual reconstruction of a portion of the tower’s wall using the database. These elements are generated the same way as standard assembly elements (bolts, screws, etc.) from the software’s built-in database. Figure 8 Knowledge database of parameterized construction elements. In Figure 9 is shown the dialog box which contains the parameters for the limestone blocks, in the same way are defined the rest of the construction elements within the database. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 83-91 88 ZSOLT BUNA et al Figure 11 Components of the wooden roof structure. Figure 9 Parameters of the limestone blocks. Using different input parameters different dimensions, angles and specific characteristics can be generated a variety of limestone blocks, like the ones in Figure 10. Figure 10 Generated limestone blocks from the knowledge database. Since during the archaeological excavations the archaeologists did not found any metal fastening elements, the engineers had to design such joints that allow the assembly of the wooden elements of the roof in such manner to support the idea that metal fastening elements were not used in the construction of this watch tower. Such joints can be seen in Figure 11. Using these methods the process of the virtual reconstruction is speeded up and in this way a very high detailed reconstitution can be created, which gives the possibility to use highly accurate engineering analyses on the created virtual models. Engineering analyses – During the excavations the archaeological data showed traces of ceramic roof tiles, and the archaeologists considered that the roof of the watch tower was covered with heat treated ceramic tiles. In order to validate the structure that can hold the roof covered with tiles the engineers had to run finite element analysis on the structure. Also because of its position on the south-eastern portion of the European continent, Romania has a temperate climate, which means at winter the precipitation is in form of snow. The authors had to take this in consideration by adding the extra weight of the snow on the roof while doing the finite element analysis. In Table 3 is shown the data used in the analysis. Table 3 Data used in the finite element analysis. Data used Value Units Weight of the tiles 26,262.5 kg Force generated by the tiles 257,372.5 N/m2 6,565.63 kg 64,343.13 N/m2 Weight of 300 mm snow 30,784 kg Force generated by the snow 301,683.2 N/m2 Weight of 300 mm snow on one side 7,696 kg Weight of the tiles on one side Force generated by the tiles on one side © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 83-91 ENGINEERING CAD TOOLS IN DIGITAL ARCHAEOLOGY Force generated by the snow on one side 75,420.8 N/m2 Total weight 57,046.5 kg Total force 559,055.7 N/m2 Total weight on one side 14,261.63 kg Total force on one side 139,763.9 N/m2 After running the finite element analysis with the information above, the translational displacement and the Von Mises stress values were identified. These values are showing where in the structure of the three-dimensional model might appear problems in a real life scenario. The translational displacement shows the displacement in centimetres which occurs under the action of forces (see Figure 12, where the roof can be seen from above). Figure 12 Display of translational displacement. The Von Mises stress shows how forces act on a structure by using different colours for different values as can be seen in Figure 13, where the structure of the roof is seen from above and where the vertical beams have to support the highest value of stress. Figure 13 Display of Von Mises stress. Since the resulted values from the finite element analysis are within the limits of safety, the validation process was a success. 89 In Table 4 can be seen the values for the translational displacement and for the Von Mises stress. Table 4 Values obtained within the FEA. Translational displacement [cm] Von Mises stress [N/m2] 0.469 5.83E+05 0.422 5.24E+05 0.375 4.66E+05 0.328 4.08E+05 0.281 3.50E+05 0.234 2.91E+05 0.188 2.33E+05 0.141 1.75E+05 0.093 1.17E+05 0.049 5.84E+04 0 1.25E+02 Validation – In the process of validation the virtually reconstructed Dacian watch tower was presented to the archaeologists and historians from the Babeș-Bolyai University, where the specialists have appreciated the methodology and were very pleased by the results. Virtual monument - After the process of validation the authors used CATIA™ V5 software platform’s built in optimization module to reduce the number of triangles which build the three-dimensional model. The raw model has 1,546,188 polygons and 629 MB in the software’s native file formats (*.CATPart, *.CATProduct). After the optimization process the model reached 58,286 polygons and 4.43 MB as an *.obj file format. In this phase were also created the additional metadata, which contain images, sound recordings, video files and text files to enrich the experience of the persons who are interested of this monument. Dissemination - After creating the virtual monument the resulted threedimensional model was introduced into an online virtual reality application (see Figure 14), where with the use of an avatar everybody can take a virtual tour in the Dacian watch tower over the internet. For © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 83-91 90 ZSOLT BUNA et al the online virtual reality application the authors chose a free hosting engine created by the same developers that created the CAD software, Dassault Systèmes, which can be accessed from www.3dvia.com. The upload of the three-dimensional models on the online platform is very intuitive and can be uploaded the native file formats from the CAD software. Figure 14 The Dacian watch tower in VR. Additional metadata was added to the existent model to enrich the experience of the visitors. The optimized model was uploaded into an augmented reality application, and also a video rendering was created featuring the virtual construction of the Dacian watch tower, where its structural composition can be seen. All these materials were presented several times for the public on different occasions, like Long Night at Museum, and archaeological sessions and the feedback of the visitors and the archaeologists was very positive. The authors recommend for VR and AR applications that the complex threedimensional models to be remodeled using the detail level that is needed for these applications, in this manner more simpler models can be created, which automatically have lower numbers of polygons and file sizes. In many cases in these environments using a simpler geometrical model with the right textures can give the same feeling as a more complex one. 5. CONCLUSIONS Using CAD software solutions in the case of destroyed ancient monuments virtual restorations, which have no documentation available about the design of the monuments, is a viable solution that can be successfully used. CAD solutions allow the use of photogrammetry, reverse engineering methods to virtually restore damaged monuments. Also a knowledge database of parameterized construction elements can be created achieving different levels of details in the restoration process. Using CAD software platforms the structure of the restored monuments can be validated using finite element analyses, creating scientifically more accurate virtual reconstructions. The achieved three-dimensional models can be optimized and introduced in different augmented reality or virtual reality applications in order to be disseminated to the general public. Depending on the level of detail that is needed for a reconstruction, the models can be crated or imported in computer graphics software, which has a more advanced renderer, and the CAD solution can be used just to validate the structure of the monument. The proposed methodology in this paper can be used by interdisciplinary teams and can be used to virtually reconstruct other types of monuments as well, due to its general character. 6. ACKNOWLEDGMENTS This paper is supported by the Sectoral Operational Programme Human Resources Development POSDRU /159/1.5/S/ 137516 financed from the European Social Fund and by the Romanian Government. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 83-91 ENGINEERING CAD TOOLS IN DIGITAL ARCHAEOLOGY 91 REFERENCES Al-kheder, S., Y. Al-shawabkeh and N. Haala (2009) Developing a documentation system for desert palaces in Jordan using 3D laser scanning and digital photogrammetry. Journal of Archaeological Science 36(2): 537-546. Amparo Núñez, A., P. Felipe Buill, M. Joaquín Regot and G. Andrés de Mesa (2012). "Generation of virtual models of cultural heritage. Journal of Cultural Heritage 13(1): 103-106. Berndt, E. and J. Carlos (2000). Cultural Heritage in the Era of Computer Graphics. Computer Graphics and Applications, IEEE 20(1): 36-37. El-Hakim, S. F., J. A. Beraldin and M. Picard (2002) Detailed 3D Reconstruction of Monuments Using Multiple Techniques. International Workshop on Scanning for Cultural Heritage, Corfu, Greece. Futragoon, N., A. Kitamoto, E. Andaroodi, M. Matini and K. Ono (2011) 3D Reconstruction of a Collapsed Historical Site from Sparse Set of Photographs and Photogrammetric Map. Computer Vision – ACCV 2010 Workshops. R. Koch and F. Huang, Springer Berlin Heidelberg. 6469: 286-295. Gelu, F. (2012) Ocolişu Mic, com. Orăştioara de Sus, jud. Hunedoara, Punct La Vămi – Poiana lui Mihu. Cronica Cercetărilor Arheologige din România. Bucharest: 91. Rua, H. and P. Alvito (2011) Living the past: 3D models, virtual reality and game engines as tools for supporting archaeology and the reconstruction of cultural heritage – the case-study of the Roman villa of Casal de Freiria. Journal of Archaeological Science 38(12): 3296-3308. Theodoropoulos, S., D. Moullou, D. Mavromati, C. Liapakis and V. Balis (2009) 3D Reconstructions of monuments and landscapes: Impressionistic or Expressionistic views of the Past? 14th International Congress "Cultural Heritage and New Technologies". Vienna: 473-480. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 83-91 Mediterranean Archaeology and Archaeometry, Vol. XX, No X, pp. 93-100 Copyright © 2014 MAA Printed in Greece. All rights reserved. VISUALIZING HIGH RESOLUTION THREEDIMENSIONAL AND TWO-DIMENSIONAL DATA OF CULTURAL HERITAGE SITES Vid Petrovic*1,2, David J. Vanoni1,2, Ashley M. Richter1,3, Thomas E. Levy1,3 and Falko Kuester1,4 1 Center of Interdisciplinary Science for Art, Architecture and Archaeology, Qualcomm Institute, University of California, San Diego 2 Department of Computer Science and Engineering, University of California, San Diego 3 Department of Anthropology, University of California, San Diego 4 Department of Structural Engineering, University of California, San Diego Received: 01/12/2013 Accepted: 02/09/204 Corresponding author: Vid Petrovic (vipetrov@ucsd.edu) ABSTRACT The combination of 3D acquisition (terrestrial and airborne LiDAR, structured light, structure-from-motion) and 2D imaging (photographic, multispectral, panoramic, orthorectified, reflectance transformation) techniques allows the geometry, appearance and other aspects of culturally significant sites to be objectively documented. Traditionally, these data are usually transformed into models such as 3D textured meshes before they are visualized or analyzed—an often time- and effort-intensive process. We propose a system for the direct visualization and analysis of such data, allowing the different aspects recorded to be layered together, and co-visualized with annotations and other relevant information. We describe the required technical foundations, including gigapoint and gigapixel visualization pipelines that enable the dynamic layering of high-resolution imagery over massive minimally-processed LiDAR point clouds that serve as the base spatial layer. In particular, we introduce the pointbuffer—a GPU-resident viewdependent point cache—as the foundation of our gigapoint pipeline, and outline the use of virtual texturing for draping of gigapixel imagery onto point clouds. Finally, we present case studies from sites in Jordan and Italy. KEYWORDS: visualization; cyber-archaeology; pointbuffer; virtual texturing; point clouds VISUALIZING HIGH-RES 3D+2D DATA OF CULTURAL HERITAGE SITES 1. INTRODUCTION Complementary to the business of elaboration and model-making, we suggest that all information, including raw, uninterpreted data, however massive, should be made available for interactive visual review and analysis. This proposed capability can also be useful in the data elaboration process itself. For example, the task of massive dataset coregistration can be made more interactive, allowing the user to verify and refine the alignment among multiple scans, models, and images in real time and at full resolution—and to spot and resolve data quality issues more rapidly. This paper discusses the on-going development of a layered visualization system that combines disparate data types such as point clouds from terrestrial laser scanning, high resolution photography and multispectral imaging in a three-dimensional environment that allows for the covisualization of annotations and other relevant information (Figure 1). In particular, we outline the gigapoint and gigapixel visualization pipelines that enable the dynamic draping of high resolution imagery over large-scale minimally processed point clouds to create a digital scaffold upon which further information can rest. Multispectral Imagery VTex Point Clouds PointBuffer … The Roman poet Horace once asked whether given the ability to render a cypress tree, one would opt to include it when commissioned to paint a sailor in the midst of a shipwreck (Horace, 19 BCE). It is a question regarding the over-use of available data: if one has the potential to create visual information- should it be made available or will its additional inclusion be distracting to one’s perception of the world around them? For virtual archaeology, there are a multitude of potential data formats that might be dynamically coalesced together to create layers of information, and which, if visualized in a navigable and dynamic format, would significantly aid in appreciating the world around us rather than detract from it. This is especially crucial for cultural heritage explorations where investigative diagnostic imaging results in a multitude of disparate data modalities that are in need of a larger visualization framework in order to experience and appreciate the contextual tapestry of information that might lead to further scientific analysis. Researchers in cultural heritage have the means of digitally recording various approximate aspects of material reality. Typically this entails first the process of surveying—of applying available digital technologies to acquire the various kinds of objective raw data about the site—followed by a process of systematic survey—of data elaboration and hands-on interpretation that results in the construction of models, maps, reports and other ‘deliverables’ useful for further study (Bianchini et al., 2012). This process of data elaboration is complex, and is both time- and labor-intensive— accordingly, there have been proposals to formalize the automation of this work (Serna et al., 2012; Pan et al., 2012). However, the challenge persists that much of the digital record remains effectively inaccessible until the products of the elaboration efforts are completed—especially for massive LiDAR scan or Structure from Motion (SfM) imaging campaigns. 94 GIS Database Live Control Figure 1 Visualization system overview. These mechanisms are intended to enable the digital documentation record to be pervasively accessible from the moment of data capture through its stages of elaboration and interpretation, and ultimately academic and public dissemination. To emphasize the utility of this system, case studies of its use on archaeological sites in southern Jordan and historical monuments in northern Italy will be presented. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 93-100 VISUALIZING HIGH-RES 3D+2D DATA OF CULTURAL HERITAGE SITES The layered system we describe is a collaborative vision between computer scientists, engineers, archaeologists, and historians. Efficient, reproducible, scientific data collection in the field is intended to be channeled into these systems and just as effectively visualized and disseminated. The following outlines the system by addressing the base layer in its data stratigraphy—point clouds—that provide the spatial and geometric foundation for other site data. 2.1 Gigapoint pointbuffer visualization with the Visualization of massive point clouds that are too large to load directly is a wellstudied problem (Wand et al., 2008; Scheiblauer et al., 2009; Scheiblauer et al., 2011; Pintus et al., 2011): a common thread is the preprocessing and reorganization of point data according to an out-of-core spatial-subdivision datastructure (commonly octree) to accelerate the selection and loading of the points relevant for visualization. To meet the needs of our project, however, we developed an alternate rapid visualization technique, based on the pointbuffer described below, that minimizes data reorganization and preprocessing requirements (relative to existing work) and gives extensive flexibility for fusing different types of data into a single visual representation. Our technique embodies a shift in strategy away from attempting to transform the input data into the form most optimal for rendering—and focusing more on ensuring adequate rendering performance even with minimally preprocessed data. It leverages the capabilities of modern graphics hardware to flexibly produce high-quality renderings that refine progressively over the course of interaction as data are streamed in from secondary storage. In contrast to most other massive-point-cloud renderers, our system requires only minimal preprocessing of data before visualization--taking time comparable to copying the dataset. The core of our gigapoint visualization pipeline is the GPU-based pointbuffer, a point-caching construct that decouples the interactive performance of visualization from the costs of the streaming large numbers of points from disk or a network server (Figure 2). data data data-points data-points selection & mapping vis-points rendering glyph pixels image sifting output GPU buffer pointbuffer 2. TECHNICAL FOUNDATIONS 95 vis-points rendering glyph pixels image Figure 2 Pointbuffer overview: the pointbuffer accumulates points for visualization, mapping input data to visualiztion points stored in a viewdependent cache The pointbuffer explicitly maintains a working set of points needed to render the view, and recycles the points that remain relevant from frame to frame—allowing greater flexibility for combining point and image data, making it possible for imagery to affect both the texture and the underlying geometry being visualized. Note that caching is a commonly applied performance optimization: a particularly fine-grained example is employed for point rendering by deferred splatting (Guennebaud et al., 2004), a generalization of deferred rendering that uses pixel-binning of point indices (references to points stored in memory buffers) to perform high-precision final point selection, and further reduces the splatting workload by re-projecting and reusing indices from the previous frame in performing visibility-splatting for the current frame. The pointbuffer we present here simplifies, generalizes, and extends this deferred splatting technique, making it possible to use point-caching as the primary mechanism for building a gigapoint visualization pipeline. We demonstrate that for large real-world datasets, effective exploitation of temporal coherence allows minimally-reorganized © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 93-100 VISUALIZING HIGH-RES 3D+2D DATA OF CULTURAL HERITAGE SITES point collections to be interactively explored, with immediate control over how the input data are mapped to visual elements. We describe some details of the pointbuffer in Section 3. storage recycle output reconstruct load shade memory pointbuffer select sift selection accumulation final image rendering Figure 3 Pointbuffer processing stages. 2.2 Gigapixel texturing To enable the use of massive imagery in visualization, we employ a virtual texturing technique (Lefebvre et al., 2004; Taibo et al., 2009; Mayer et al., 2011), a combination of classical MIP-mapping and virtual memory approaches. The approach is to store images in a tiled multi-resolution pyramid, and load only the tiles needed to perform the texture lookups requested. For a given view, the set of tiles corresponding to the observed texture lookups (2D+levelof-detail/scale coordinates) is determined. In our variant of virtual texturing, this step is performed entirely on the GPU, with only the results being read back by the CPU— permitting a tight coupling with the gigapoint pipeline. The needed tiles are then fetched from disk and uploaded to the GPU tile cache. We keep the peak of the image pyramid always resident so that all lookups can be satisfied (albeit at a lower resolution) even before the more appropriate are uploaded, and perform trilinear filtering between the two bracketing levelsof-detail once the corresponding tiles are cached. 3. THE POINTBUFFER The pointbuffer is a GPU-based datastructure that decouples the loading and rendering pipeline stages by buffering the visualization points needed for rendering. Vispoints remain in the pointbuffer for as long as they are appropriate for the view 96 being rendered. Whenever the view setup changes—due to camera motion, or a change in selection, visualization or mapping parameters—the pointbuffer keeps the vispoints that are still appropriate for the new view. The pointbuffer offers two high-level operations: sift and output. The sift operation processes a datapoint batch, adding to the buffer the visualization points appropriate for the view, and reporting as feedback the number of points added. The output operation streams out, in part or in full, the buffered vispoints for further processing or rendering. The points in the buffer are recycled between frames, accounting for any change in view. This can be performed by ping-ponging between two buffers: points are output from the old buffer and sifted into the new buffer. A sifting cycle consists of two steps: 1. Recycle points already in the buffer. 2. Sift in batches of new points. Points accumulate in the pointbuffer over many frames, allowing the resulting renderings to refine gradually over time, even 3.1 Pointbuffer binning Points in the pointbuffer are stored in a grid of bins, with at most one point per bin, with the binning scheme determining the mapping of image point samples to image pixels. In the simplest scheme, the binning grid coincides with the image pixel grid, allowing the pointbuffer to hold at most one vispoint per image pixel. More elaborate binning schemes are preferable in practice, with multiple bins per pixel, or a nonuniform distribution of bins, such that there are more samples per pixel near the center of the image than elsewhere, for example. The binning scheme and capacity— the number of bins—can be set independently of the output image size, and is a parameter that influences performance and attainable rendering quality. In practice, image-center-biased bin distributions provide the best quality-performance bal- © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 93-100 VISUALIZING HIGH-RES 3D+2D DATA OF CULTURAL HERITAGE SITES ance, with oversampling near the image center (for crisp, low-alias detail rendition) and a manageable total bin count (for performance). 3.2 Point selection and loading For this work, we understand a point dataset to be a collection of distinct point sets, or point clusters. We ensure through preprocessing (described below) that each cluster is stored as a sequence of points such that: 1. any contiguous subsequence of a cluster is expected to have approximately the same spatial distribution; and 2. different clusters ideally have different spatial distributions. The first property above allows us to use a contiguous subsequence of cluster points, such as a prefix sequence, to render an approximation for the cluster. It also makes it possible for us to estimate the contribution that points from a cluster will make to a final rendered image by measuring the contribution made to an approximate image by points from a smaller prefix sequence. The second property differentiates the clusters from a selection standpoint, allowing us to make gains in loading efficiency by view-dependently favoring selection from some clusters over others. The contribution that the points from a single cluster make to the rendered image is view-dependent. We quantify this cluster contribution as the number of points binned from the cluster. For any given view, clusters in general have differing levels of contribution. For close-up views, points from relatively few clusters tend to dominate, with most clusters contributing few or no points. For total views, in which most of the scene is visible, it can happen that many or all clusters contribute relatively few points each. The selection engine focuses the loading and sifting effort on the clusters that contribute the most to the image. Initially, the clusters are assumed to make contributions proportional to their point counts. As 97 points are sifted into the pointbuffer, the feedback returned—the number of points accepted into the pointbuffer—is used to estimate each cluster’s contribution, and to guide the selection process over the following sifting cycles. 3.3 Dataset preprocessing The purpose of preprocessing is to generate and store the datapoint clusters in a form suitable for selection. Many kinds of point datasets are already stored, or can be efficiently exported, as sequences of datapoints exhibiting at least some spatial coherence. Examples include LiDAR data (stored in scan order), volume data (stored in slices or blocks), image data (e.g., SfM), and other spatially-organized datasets. Transcoding such a dataset is particularly simple and fast, since the data can be broken up into clusters sequentially. The order of points in each cluster is then randomized, yielding a point cluster with the desired characteristics. The preprocessing procedure is: 1. Fill an array of the desired (cluster) size sequentially with datapoints, optionally transforming or reformatting the points. 2. Shuffle the points in the array. 3. Store the array as a cluster. 4. Repeat steps 1-3 until all points have been processed. Note that the transcoding time is linear in the number of points. Since the points in each cluster are shuffled, any prefix sequence for the cluster is a random subset of the full point set, and will therefore on average have approximately the same spatial distribution as the full set. 4. PERFORMANCE CHARACTERISTICS On a midrange laptop (ca. 2011, Intel Core i5-254M, 4GiB RAM, 2.6GHz, AMD Radeon HD 6750M with 512MiB graphics RAM), the gigapoint pipeline maintains rates above 30Hz at 1280x720 (for all datasets tested, with largest at ~2.75 billion points), and the virtual texturing pipeline © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 93-100 VISUALIZING HIGH-RES 3D+2D DATA OF CULTURAL HERITAGE SITES maintains rates above 60Hz at 1920x1080 (tested with synthetic data for ~16 billion pixels). The combined pipelines have performance similar to that of the point pipeline alone. Renderings usually refine over the course of a few seconds at most for point rendering, and well under a second for texturing. 4.1 Preprocessing performance Preprocessing time is dominated by the cost of reading the raw data and writing the output clusters: the point reformat- ting and shuffling costs are generally dwarfed by I/O costs. The preprocessing rate depends upon the input data format, the size of the output point format, and the speed of the source drive (and destination drive, if different). The rate ranges from 0.5 million points per second for transcoding text PTX files using a single external USB drive to several million points per second with high-performance drives. The overhead relative to simply reading from the source and writing to the destination is in our implementation caused by input parsing inefficiencies and insufficient I/O overlapping. We expect that further implementation optimizations can bring the transcoding costs down to near that of a direct copy operation. 98 overall, refinement rarely takes more than a few seconds. 5. CASE STUDIES 5.1 The Byzantine Church at Petra The current test-case for our system is to combine terrestrial laser scanning point clouds with systematic high resolution photography, diagnostic maps, and corresponding semantic information collected during and for the 2012 diagnostic imaging survey of Petra's Byzantine Mosaic Church (see Figures 4 and 5) at the behest of the American Center of Oriental Research's Temple of the Winged Lion and Environs Conservation Projects (Levy et al., 2013). In this case, our work is aimed at creating not just a digital tourist tool for the Petra UNESCO World Cultural Heritage Site in Jordan, but in building a visualization tool that can be useful for the on-going research and conservation monitoring at the site. 4.2 Selection performance The loading of cluster blocks is in practice done at the disk’s maximum read rate, or at a user-specified lower rate. The amount of time it takes for a given view to fully refine depends on how different the view is from recent views and the level of main-memory caching that has been attained. As the rendering refines, the progress that is being made is visually evident, and this fact tends to make subjectively acceptable even longer waits than are encountered in practice. The refinement time is the greatest after a cold start, when all data must be loaded from disk. The system tends to reach steady-state after less than a minute of interaction (usually in under 30 seconds); Figure 4 The Byzantine Church at Petra, Jordan: LiDAR of mosaic. Figure 5 The Byzantine Church at Petra: LiDAR of mosaic with photographic overlay. Acquisition of data for the Byzantine Mosaic Church was a field trial of rescue © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 93-100 VISUALIZING HIGH-RES 3D+2D DATA OF CULTURAL HERITAGE SITES archaeology methodologies under development, and was performed in under two hours. Four researchers (one LiDAR operator, two photographers, and one assistant) worked concurrently to maximize coverage of the site within the window of time available. Over 3000 digital photographs were taken in a pattern suitable for SfM camera pose estimation and 3D modeling. The images were subsequently processed using Agisoft PhotoScan (a commercial SfM/photogrammetry software package) to produce high-resolution (32kx32k pixel) orthophotos of the floor mosaics, which were then draped dynamically onto LiDAR data using the system described in this paper. Note that it is also possible to drape images individually onto the point cloud using the camera parameters (intrinsic and pose) estimated by PhotoScene, bypassing the generation of orthophotos. 5.2 Salone dei Cinquecento, Palazzo Vecchio Our second example is the diagnostic imaging study of the Hall of Five-hundred at Palazzo Vecchio, in Florence, Italy. This Renaissance site is the possible location of a Leonardo Da Vinci painting (the Battle of Anghiari) whose whereabouts have been unknown for over 450 years. Over one billion LiDAR points were collected, with 1mm-resolution capturing the six frescoes and eight sculptures along the east and west walls. Additionally, high-resolution (~30kx20k pixel) visible-light imagery was captured by a Panoscan rotating line-scan camera, as well as a thermal image mosa- 99 ics. In Figure 6, the LiDAR data is shown layered with the Panoscan imager and thermal annotations (in green). A preliminary version of the system outlined in this paper (with additional layering of CAD models of the Palazzo and groundpenetrating-radar imagery of the wall and Vasari fresco) was used to help guide an endoscopic study in 2011. Figure 6 Battaglia di Anghiari project: highresolution visible-light image captured by a Panoscan rotating line-scan camera, and a thermal image mosaic (as a green overlay), on top of a LiDAR point cloud. 6. CONCLUSIONS We have described a small set of mechanisms sufficient for dynamically visualizing minimally processed 3D point clouds along with high-resolution 2D imagery. The approaches presented—pointbufferbased gigapoint and virtual-texturingbased gigapixel pipelines—allow massive digital documentation datasets, comprising LiDAR and multispectral photography, to be viewed and inspected interactively throughout their digital lives, from acquisition to elaboration to dissemination. REFERENCES Bianchini, C., Borgogni, F., Ippolito, A., Senatore, L.J., Capiato, E., Capocefalo, C. and Cosentino, F. (2012) From surveying to representation theoretical background, practical issues, possible guidelines. Virtual Systems and Multimedia 2012. Guennebaud, G., Barthe, L. and Paulin, M. (2014) Deferred splatting. Computer Graphics Forum, 23(3):653–660, Sept. 2004. Horace, Ars Poetica 21. Lefebvre, S., Darbon, J. and Neyret, F. (2004) Unified texture management for arbitrary meshes. Research Report 5210, INRIA. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 93-100 VISUALIZING HIGH-RES 3D+2D DATA OF CULTURAL HERITAGE SITES 100 Levy, T.E., Tuttle, C.A., Vincent, M., Howland, M., Richter, A., Petrovic, V. and Vanoni, D. (2013) The 2012 Petra Cyber-Archaeology Cultural Conservation Expedition: Temple of the Winged Lions and Environs, Jordan. Antiquity (Project Gallery), http://antiquity.ac.uk/projgall/levy335. Mayer, I., Scheiblauer, C. and Mayer, A.J. (2011) Virtual texturing in the documentation of cultural heritage the Domitilla catacomb in Rome, Proc. Of XXIIIrd International CIPA Symposium. Pan, X., Schiffer, T., Schrottner, M., Havemann, S., Hecher, M., Berndt, R. and Fellner, D.W. (2012) A scalable repository infrastructure for CH digital object management. Virtual Systems and Multimedia 2012. Pintus, R., Gobbetti, E. and Agus, M. (2011) Real-time rendering of massive unstructured raw point clouds using screen-space operators. The 12th International Symposium on Virtual Reality, Archaeology and Cultural Heritage VAST. Scheiblauer, C., Zimmermann, N. and Wimmer, M. (2009) Interactive Domitilla Catacomb exploration. The 10th International Symposium on Virtual Reality, Archaeology and Cultural Heritage VAST. Scheiblauer, C. and Wimmer, M. (2011) Out-of-core selection and editing of huge point clouds. Computers & Graphics, 35(2):342-351. Serna, S.P., Schmedt, H., Ritz, M. and Stork, A. (2012) Interactive semantic enrichment of 3D cultural heritage collections. The 13th International Symposium on Virtual Reality, Archaeology and Cultural Heritage VAST. Taibo, J., Seoane, A. and Hernández, L.A. (2009) Dynamic virtual textures. Journal of WSCG, Volume 17, Number 1. Wand, M., Berner, A., Bokeloh, M., Jenke, P., Fleck, A., Hoffmann, M., Maier, B., Staneker, D., Schilling, A. and Seidel, H. Processing and interactive editing of huge point clouds from 3D scanners. Computers & Graphics, 32(2):204-220. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 93-100 Mediterranean Archaeology and Archaeometry, Vol. 14, No 4, pp. 101-108 Copyright © 2014 MAA Printed in Greece. All rights reserved. PHOTOGRAMMETRY IN THE FIELD: DOCUMENTING, RECORDING, AND PRESENTING ARCHAEOLOGY Matthew D. Howland1, 2, Falko Kuester2 and Thomas E. Levy1 1University of California, San Diego, Department of Anthropology, USA of California, San Diego, Qualcomm Institute, USA 2University Received: 03/12/2013 Accepted: 17/06/2014 Corresponding author: Matthew D. Howland (mdh5169@gmail.com) ABSTRACT The development of three-dimensional documentation technologies such as LiDAR and Structure from Motion (essentially digital photogrammetry) has led to a recording revolution, as these methods are increasingly applied to field archaeology. 3D methods have the potential to become an integral part of the archaeological toolkit, as they have the capability to produce spatially-referenced outputs, such as orthophotos and digital elevation models (DEMs), with greater efficiency than traditional methods. The combination of Structure from Motion and low-altitude aerial photography can facilitate the production of these GIS outputs, which can then be used for digitization or as basemaps. These methods allow for accurate and precise recording with a relative minimum of field time. As the existing body of 3D data increases in size, museums have the unique opportunity to be able to take advantage of these datasets to update their exhibits and display archaeological context and the process of excavation through visualizations of 3D models. The spread of 3D documentation and recording in archaeology may provide a unique opportunity for collaboration between these two professions, and allow for archaeology to improve its public outreach. The methodology presented here is based on field research in Jordan KEYWORDS: Cyber-archaeology, Photogrammetry, Low-altitude Aerial Photography, Jordan 102 1. INTRODUCTION The development of new techniques of archaeological field recording has the potential to revolutionize the documentation, interpretation, and presentation of archaeological sites, and the archaeological process. Three dimensional recording techniques, especially Structure from Motion, allow for new methods of gathering data from archaeological sites, in ways that are more efficient, precise, and accurate than traditional methods. The data acquired using these approaches can also provide the basis for new ways of displaying archaeological context in a museum setting. 2. BACKGROUND For archaeologists, context is the key to information about artifacts and sites. Recording provenience and context in the field is essential to preserving the information uncovered by excavation, given that field archaeology is an inherently destructive endeavor (Wheeler 1954). Excavators and surveyors remove artifacts from their contexts, remove, discard, and mix up the soil and stone that surround the artifact, and even annihilate potentially precious sources of data that are either too expensive, impractical, or, as of now, impossible to study. Once a site is excavated and its remains are moved to the laboratory, the potential research value of what is taken is reduced to the limitations set by the records and recovered artifacts of the excavator. Spatial relationships of artifacts and loci at the site become impossible to investigate beyond what is documented in the field. Mortimer Wheeler, an iconic advocate of detailed recording, expounded upon the need for recording of archaeological strata (an essential variable in context)— and criticized some of his contemporaries for their failure to do – as early as the 1950s (Wheeler 1954). Since these times, standards of archaeological recording have improved dramatically. The modern archaeologist justifies his destruction of cultural heritage through extensive documentation of both artifact and context. The tension HOWLAND et al created by the elimination of data at the hands of one who studies it is at least partially eased by the archaeologist’s attempt to preserve the context of discovery in a number of different ways. By recording information that can be used to recreate the circumstances of the field, archaeologists facilitate the efforts of later scholars to reinterpret their data and also improve their own ability to provide more concrete evidence for their assertions. Recent developments in technology have allowed for an improvement in field recording techniques to the point where it is now possible to digitally recreate the circumstances of excavation in the lab (Levy 2013). This is made possible by the availability of technologies such as LiDAR and Structure from Motion, which have moved the possibilities of recording and presenting archaeological context into the third dimension. The potential to digitally create photorealistic and spatially accurate representations of objects or areas of interest has opened up a new realm of documentation – that of 3D recording. Archaeological projects have made increasing efforts to capitalize on this development (Lambers et al 2007; Lerma et al 2010; Ortiz Sanz et al 2010; Al-kheder et al 2010; Olson et al 2013; Verhoeven et al 2012). Meanwhile, laser scanning and photogrammetry have been already widely applied to the documentation of ancient monuments for conservation purposes and in museums, mostly in digitizing artifacts, whether for the purposes of documentation or digital display (Wachowiak and Karas 2009; Bruno et al 2010; Yilmaz et al 2007; Pavlidis et al 2007). Attempts to use some of the increasingly available 3D datasets from archaeological projects for museum display of the contexts artifacts are recovered from are less common. Techniques of three-dimensional recording of archaeological sites allow for the creation of a fully-three-dimensional record of archaeological excavation with high temporal and spatial resolution (Olson et al 2013). The acquisition of this type of data also potentially allows for new approaches © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 101-108 PHOTOGRAMMETRY IN THE FIELD to presentation of archaeology to the public. 3. METHODS The 2012 Edom Lowlands Regional Archaeology Project (ELRAP) provides a case study of one method of data collection well-suited to creating a photorealistic, three-dimensional record of an archaeological site. Members of the UC San Diego ELRAP team conducted intensive excavation at a number of sites in southern Jordan’s Wadi Arabah, among which was Wadi Fidan 61, a Neolithic period site. The team used a 1-ply Kingfisher Aerostat K14U-SC balloon, with dimensions of ca. 3.6 m x 3.0 m, volume of ca. 21.0 m3, and lift of ca. 13.6 kg when fully inflated. The balloon was tethered to a reel by 800-lb. strength Spectra fiber line and manipulated by a ground-based operator. In order to perform low-altitude aerial photography, the balloon was outfitted with a custom triangular frame capable of holding two high-resolution (15.1 megapixel) Canon EOS 50D Digital Single-Lens Reflex (DSLR) cameras equipped with 18mm lenses. These DSLRs were also applied independently of the balloon rig to record the site in a number of different ways and resolutions. Needing to document an excavated tomb of ca. 2x3 m and the entire site at which the tomb is located (ca. 6 ha.), ELRAP team members developed custom strategies designed to record every feature with overlapping high-resolution images. Photographic data collection was oriented towards creating high-quality 3D models using Structure from Motion technology, a digital procedure that updates and uses traditional photogrammetric techniques to create a three-dimensional model from points of similarity between photographs of the same object taken from different angles. To that end, the team applied custom, individualized strategies to the tomb and to the site. To document the tomb, ELRAP personnel captured photographs of the subject from the ground, attempting to achieve 360 degrees of over- 103 lapping (by ca. 80%) photographic coverage in order to ensure that every aspect of both the tomb and the excavated square would be recorded by several photographs. The site itself, consisting of a 6 hectare prominence rising from the Wadi Fidan, required more expansive coverage, necessitating the use of the balloon photography system briefly described above. By maneuvering the balloon in transects over the mound, the ELRAP team acquired highresolution and overlapping aerial imagery covering nearly the entire site. The transects were designed with an intent to collect photographs with an ideal overlap of 50% between adjacent images, both along and between transects. The data needed for the recording of the tomb (ca. 70 photos) took only five minutes, in the field. Field recording of the entire site using the balloon (ca. 300 photographs) was collected in approximately 2 hours in the field, a rapid pace allowing for the creation of a 3D model of the site within a single day. The same system of recording was used on excavation units at other sites in the ELRAP campaign twice each day, with a needed downtime from excavation of only ca. 20 minutes in the morning prior to the start of the day’s work. With photographic data of areas of interest collected and appropriately sorted and stored within the ELRAP database, the team applied the commercially-available Agisoft Photoscan program (run on a desktop PC outfitted with an 8 core 3.07 GHz Intel Core i7 processor, 12 GB of RAM, a 1 TB HD, and an NVIDIA GeForce GTX 580 GPU) to the photographic dataset in order to create 3D models of recorded areas. From this three-dimensional dataset, the team also produced 2D orthophoto and DEM outputs. Agisoft Photoscan is a userfriendly software package providing a comprehensive Structure from Motion approach, with the ability to process unsorted photographs into a photorealistic, geometrically-accurate, and georeferenced 3D model. The process of generating 3D models through this program’s workflow breaks down into three main steps. After © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 101-108 104 uploading a set of photos to the program, the first processing stage is known within the program as “Align Photos.” During this step, the unsorted dataset is developed into a point cloud representing the points of similarity between the different images (with the identification of points known generally as the stereo-matching problem). As part of this process, the location of where each photograph was taken is calculated using the angles of capture of each image and a process called photogrammetric bundle adjustment (Triggs et al 2010). This stage of model development is relatively computationally intensive, taking 1-3 hours to process on the ELRAP computer, depending on the number of images used to create the model. The output of this stage of processing consists of a lowdensity point cloud (usually of ca. 100300,000 points), which can be edited, cropped, and cleaned up to both facilitate and improve the accuracy of the ensuing processing stages. The point cloud is used as the basis for the next step of processing, which is known as “Build Geometry.” This phase produces a solid geometrical model based on the point cloud, with the possibility to set standards of model accuracy and resolution (in the form of the number of geometric faces of the model) determining the quality of the output of this stage. This model can also be trimmed and cropped according to areas of interest and/or quality. Building the geometry of the model is also fairly computationally intensive, taking 1-3 hours, the total of which also depends on the size of the input dataset. The final step of model processing is known as “Build Texture,” which consists of overlaying the original images used to create the model back onto the geometric form created in the prior stages. The way in which the images are overlaid can be customized according to the desired function and appearance of the model (See Verhoeven 2011 for more information on Agisoft Photoscan workflow). This final stage of model processing is the quickest and least-intensive, taking only ca. 5-30 minutes. This workflow was applied to each of the areas of HOWLAND et al interest, with care taken to maximize the resolution and quality of the developing model at each step of processing. The tomb model took approximately 2.5 hours to develop, the excavation unit requiring ca. 4 hours, and the site model – with the largest input dataset – processing in ca. 7 hours. The last aspect of the workflow taking place within Agisoft Photoscan consisted of georeferencing the models, using control points recorded in the field with a total station or GPS unit. The ELRAP team manually entered the coordinates of each point into the program for each area of interest, thereby georeferencing the models and preparing them for the exporting of data in GIS-compatible formats. 4. RESULTS The workflow outlined above was primarily designed with the intention of producing spatially-referenced twodimensional outputs with the highest degree of accuracy and precision possible. Developing a 3D model allows for the creation of orthophotographs, top-down images that are corrected for lens and elevation distortion (Lo 1974), and digital elevation models (DEMs). Orthophotos provide a more accurate basis for digitization of architectural features than do georeferenced photos (Verhoeven et al 2012) and are a critical part of the ELRAP campaign, which produces daily top plans of extant architecture. These plans are crucial for interpretation and publication of the site, and their accuracy – facilitated by their basis in orthophotos – is of the utmost importance. Digital elevation models also provide a useful basemap for contextualizing sites, and can be used to create contour lines within ArcGIS or other GIS software. Structure from Motion allows for the rapid (ca. 10 hours) production of high-resolution, precise (ca. 5-10 cm) DEMs that would otherwise take days or weeks of valuable field time to produce with traditional methods of EDM survey, requiring hundreds or thousands of points (cf. Louhaichi et al 2003). The documentation at WF61 pro- © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 101-108 PHOTOGRAMMETRY IN THE FIELD duced a 2cm resolution orthophoto of the entire site (with submeter spatial accuracy across the mound) (Figure 1). 105 Finally, a high-resolution orthophoto of the tomb (with ca. 2 cm spatial accuracy) was created for the purposes of digitization of the rock features of the tomb with the highest degree of accuracy possible (Figure 3). Figure 1 Orthophoto of WF61 with 2cm resolution. This image, corrected for lens and elevation distortion, was exported from a 3D model of the site, allowing for the creation of the orthophoto. The team also manufactured a 4cm resolution DEM (Figure 2) suitable for the production of quality contours. Figure 2 DEM of WF61 with 4cm resolution. This model was exported from a 3D model of the site, allowing for the creation of the DEM. Figure 2 Orthophoto of tomb at WF61 with high resolution. This orthophoto was used as the basis for digitization of the rocks surrounding the tomb. 5. DISCUSSION The models produced in the process of generating these 2D georeferenced outputs for the ELRAP GIS represent a spatiallycomprehensive 3D record of the site. With models both extensive enough to contain data for the entire site and intensive enough to show specific archaeological contexts, the three-dimensional documentation of excavation and research at Wadi Fidan 61 is relatively complete. The speed of recording, discussed above, using the combined tactics of balloon photography and Structure from Motion have allowed the ELRAP team to create a record with high spatial and temporal resolution without sacrificing valuable time in the field. The temporal efficiency of this workflow was made evident to ELRAP team members by an internal comparison between the extent of data collection coverage attained by the previously outlined photographic/Structure from Motion workflow (6 ha. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 101-108 106 sites recorded in ca. 2 hours of fieldwork) and that of ELRAP terrestrial laser scanning efforts, which required similar timeframes for recording only small subsections of sites, given the necessity for multiple overlapping laser scans. A comparison between the relative accuracy and precision of these Structure from Motion and terrestrial laser scanning is beyond the scope of this paper, although we suggest that Agisoft Photoscan-developed models and GIS products are well within the limits of acceptable spatial error for archaeological purposes. This case study shows that even relatively technologically-intensive methods of 3D recording have the potential become an integral part of archaeological field projects such as ELRAP, given the efficiency of photogrammetric field recording at multiple scales. With Structure from Motion-based approaches, archaeologists can acquire three-dimensional datasets with high temporal and spatial resolution and accuracy with minimal to no disruption to other avenues of field investigation. Equally as important to the archaeologist is that the combination of low-altitude aerial photography and Structure from Motion allows for the creation of GIS-based data that are unique in their combination of extent and resolution, as compared to the products of other methods. Sitewide orthophotographs of ca. 2 cm resolution are both substantially more detailed than satellite imagery and also free of the lens and elevation distortion inherent to traditional vertical photography. Meanwhile, site DEMs of ca. 5 cm resolution provide a high-precision basis for the creation of site contours or elevation-based spatial analyses. Compared to satellite-acquired elevation data – often available at resolutions no better than 30 meters, larger than many small archaeological sites – a resolution allowing researchers to pick out even small features of sites on a DEM is extremely valuable. ELRAP’s development of a 3D recording system to produce GIS content has also resulted in the creation of a great deal of aes- HOWLAND et al thetically-pleasing and accurate 3D data showing archaeological context. We believe that these datasets – already produced for reasons of documentation and preservation – can be used in a museum with a relative minimum of effort and the reward of being able to display archaeological context as it was seen by the excavators during the process of investigating the site. 3D models created primarily for purposes of archaeological conservation and documentation can have a secondary use in presentation of archaeology, in digital museums or in traditional museums with television or computer displays (Bruno et al 2010; Carrozzino and Bergamasco 2010). The possibility of acquiring these high-quality and aesthetically-pleasing datasets has the potential to be a boon for museums, which now have the possibility of displaying these models in a number of different ways. The usefulness of this type of context-focused display was demonstrated by a cyberarchaeology exhibit titled “EX3: Exodus, Cyber-Archaeology and the Future” developed and presented by the University of California, San Diego’s Qualcomm Institute (Salamon et al in press) (Figure 5). Figure 3 An undergraduate docent presenting the animated 3D model of WF61 to an interested visitor, pointing out the tomb in the context of the larger site. Given that context is an essential variable in archaeology, it seems also important to present this data in a museum setting. We propose that three-dimensional data is an © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 101-108 PHOTOGRAMMETRY IN THE FIELD ideal and cost-efficient solution to the problem of how to display 6. CONCLUSION The photogrammetric method of Structure from Motion is a viable tool for archaeological documentation and field recording. Structure from Motion-based approaches are practical and useful on archaeological projects given the efficiency of field collection and the precision and accuracy of the datasets produced using these techniques. We suggest that photogram- 107 metric methods in combination with both ground-based and low-altitude aerial photography represents an effective workflow for 3D documentation and collection of spatial data at scales ranging from small excavation units to entire sites. Additionally, the increasing availability of 3D data at high spatial and temporal resolution provides museums with the opportunity to present the process of excavation and archaeological context to their publics, expanding the possibilities of public outreach for archaeology. REFERENCES Al-Kheder, S., Al-shawabke, Y., and Haala, N. (2009) Developing a documentation system for desert palaces in Jordan using 3D scanning and digital photogrammetry. Journal of Archaeological Science, vol. 36, 537–546. Bruno, F., Bruno, S., De Sensi, G., Luchi, M., Mancuso, S., and Muzzupappa, M. (2010) From 3d reconstruction to virtual reality: A complete methodology for digital archaeological exhibition. Journal of Cultural Heritage, vol. 11, 42–49. Carrozzino, M., and Bergamasco, M. (2010) Beyond virtual museums: Experiencing immersive virtual reality in real museums. Journal of Cultural Heritage, vol. 11, 452– 458. Lambers, K., Eisenbeiss, H., Sauerbier, M., Kupferschmidt, D., Gaisecker, T., Sotoodeh, S., and Hanusch, T. (2007) Combining photogrammetry and laser scanning for the recording and modeling of the Late Intermediate period site of Pinchango Alto, Palpa, Peru. Journal of Archaeological Science, vol. 34, 1702–1712. Lerma, J.L., Navarro, S., Cabrelles, M., and Villaverde, V. (2010) Terrestrial laser scanning and close range photogrammetry for 3D archaeological documentation: the Upper Palaeolithic Cave of Parpallo´ as a case study. Journal of Archaeological Science, vol. 37, 499–507. Levy TE. 2013. Cyber-Archaeology and World Cultural Heritage: Insights from the Holy Land. Bulletin of the American Academy of Arts & Sciences LXVI:26-33. Lo, C.P. (1973) The Use of Orthophotographs in Urban Planning. The Town Planning Review, vol. 44, 71–87. Louhaichi, M., Borman, M.M., Johnson, A.L., and Johnson, D.E. (2003) Creating low-cost high-resolution digital elevation models. Journal of Range Management, vol. 56, 92–96. Ortiz Sanz, J., de la Luz Gil Docampo, M., Martinez Rodriguez, S., Terese Rergo Sanmartin, M., and Mejide Cameselle, G. (2010) A simple methodology for recording petroglyphs using low-cost digital image correlation photogrammetry and consumer-grade digital cameras. Journal of Archaeological Science, vol. 37, 3158–3169. Olson, B.R., Placchetti, R., Quartermaine, J., and Killebrew, A.E. (2013) The Tel Akko Total Archaeology Project (Akko, Israel): Assessing the suitability of multi-scale 3D field recording in archaeology. Journal of Field Archaeology, vol. 38, 244–262. Pavlidis, G., Koutsoudis, A., Arnaoutoglou, F., Tsioukas, V., and Chamzas, C. (2007) Methods for 3D digitization of Cultural Heritage. Journal of Cultural Heritage, vol. 8, 93–98. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 101-108 108 HOWLAND et al Pearce, S.M. (1993) Museums, Objects, and Collections Washington, D.C., Smithsonian Institution Press. Salamon A, Ward S, McCoy F, Hall J, and Levy TE. in press. Inspired by a Tsunami? Earth Sciences Perspectives of the Exodus Narrative. In: Levy TE, Schneider T, and Propp WHC, editors. Israel's Exodus in Transdisciplinary Perspective - Text, Archaeology, Culture and Geoscience. New York: Springer. Sylaiou, S., Liarokapis, F., Kotsakis, K., and Patias, P. (2009) Virtual Museums, a Survey and Some Issues for Consideration. Journal of Cultural Heritage vol. 10, 520–528. Triggs, B., McLauchlan, P.F., Hartley, R.I., and Fitzgibbon, A.W. (2000) Bundle adjustment – A modern synthesis. In Vision Algorithms: Theory and Practice, Springer-Verlag, Berlin, Germany. Verhoeven, G. (2011) Taking computer vision aloft – Archaeological three-dimensional reconstructions from aerial photographs with PhotoScan. Archaeological Prospection, vol. 18, 67–73. Verhoeven, G., Doneus, M., Briese, Ch., and Vermeulen, F. (2012) Mapping by matching: a computer vision-based approach to fast and accurate georeferencing of archaeological aerial photographs. Journal of Archaeological Science, vol. 39, 2060–2070. Wachowiak, M.J., and Karas, B.V. (2009) 3D scanning and replication for museum and cultural heritage applications. Journal of the American Institute for Conservation, vol. 48, 141–158. Wheeler, M. (1954) Archaeology From the Earth Oxford, Clarendon Press. Yilmaz, H.M., Yakar, M., Gulec, S.A., and Dulgerler, O.N. (2007) Importance of digital close-range photogrammetry in documentation of cultural heritage. Journal of Cultural Heritage, vol. 8, 428–433. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 101-108 Mediterranean Archaeology and Archaeometry, Vol.14, No 4, pp. 109-116 Copyright © 2014 MAA Printed in Greece. All rights reserved. OPENDIG: CONTEXTUALIZING THE PAST FROM THE FIELD TO THE WEB Matthew L. Vincent*1,2, Falko Kuester2 and Thomas E. Levy1,2 1Levantine and Cyber-Archaeology Laboratory, University of California, San Diego 2Qualcomm Institute, University of California, San Diego Received: 20/01/2014 Accepted: 08/05/2014 Corresponding author: Matthew L. Vincent (mlvincent@ucsd.edu) ABSTRACT Data recording is one of the primary requirements of any archaeological project. Some projects rely on the traditional pen-and-paper methods, while others have begun to employ field data recording applications through mobile computing platforms. The former method relies on later transcription of the data, while the later passes over this step, integrating the data from various devices at some later point. Many rely on commercial solutions to solve their data recording needs. Well-known platforms, which have had a long and successful track record with databases, are now being employed for archaeological databases. Although these robust platforms provide straightforward solutions, they are expensive and not easily extensible. OpenDig was developed with a focus on open source frameworks, with the idea that future expansion would be important for any archaeological database. By utilizing open source tools that were born in the World Wide Web, OpenDig provides a complete framework for archaeological data from the field and post-excavation studies. The three main tools that make up the OpenDig framework are: 1) a field recording application for describing archaeological contexts, associated photos, geospatial data, and find; 2) a lightweight data reader and editor for deployment in field laboratories; 3) a full web application for a more complete tool set for reviewing, analysing and disseminating these data acquired from the field. Three tools, on their own, may not seem very different from other solutions available to archaeologists today. However, OpenDig demonstrates the viability of using open source tools and open source data to create a complete system for data recording, analysis and dissemination. The future of archaeological data lays in finding ways to link disparate data sets from various projects and being able to make sensible comparisons. This can only be achieved by providing open access to these data and creating common interfaces that allow archaeologists to link their data with others. KEYWORDS: OpenDig, data recording, field archaeology, Cyber-archaeology, archaeology cyber-infrastructure 110 VINCENT et al 1. INTRODUCTION As archaeology is the ‘science of destruction,’ it depends on careful and meticulous data recording in order to preserve the archaeological contexts where artifacts are found for analyses and modeling. These data become the metadata that describe the context of artifacts, samples and geographic data. Yet, for many projects, the recording of these data takes a secondary place in the field as well as the publication process. Data recording is often done on paper forms that later must be digitized, which is often a tedious process that ends up being neglected. Others have adopted digital field recording methods, but have not found ways to publish them to the Web – an essential process to share and disseminate data. Instead, these data are disseminated to the public only through the lens of interpretation in the form of journal articles, books and other written publications. Although these interpretations are still some of the most important methods for disseminating this information, the primary data must find a place in the publication process, allowing others to use these data as part of their research as well. This paper looks at OpenDig, a framework for archaeological meta-data, and the role it plays in the recording, analysis and dissemination of archaeological data. Furthermore, it seeks to encourage the rapid publication of primary data through machine-readable formats, opening up new avenues of research through open data. Finally, this article encourages an agile approach to archaeological data. Rather than trying to conform disparate archaeological datasets to a single schema, ways should be found to link datasets using Web-based standards proven in the field today. 2. OPENDIG: AN ARHCAEOLOGICAL FRAMEWORK OpenDig has its foundations in the Madaba Plains Project, namely Tell al'Umayri, where a recording system was developed with a focus on holistic verbal descriptions that attempt to detail the archaeological context as much as possible (Brower 1989, Clark 2011). The forms include detailed information that describe the type of sediment, building style, related colors and other information (see figure). Unfortunately, such information does not readily lend itself to tabular data, like that found in the widespread Structured Query Language (SQL) databases popular today. These databases rely on tables with rows and columns for data storage. Rows are related through common identifiers, allowing the user to compile together rows from various tables to connect data together. The problem presented by these forms was finding a way to correctly represent them in an electronic format, particularly when it required breaking a single document apart into multiple tables. Figure 1 An example of the Madaba Plains Project paper based recording system. 2.1 Schemaless Data Apache's CouchDB (Anderson, Lehnardt, and Slater 2010) is one of the popular NoSQL databases currently available. NoSQL databases, as the name suggests, move away from tabular recording systems and instead store data as documents. The flexible database schemas make it ideal for archaeological recording, as small variances in recording methodologies from site to site make archaeological databases difficult to implement across a varie- © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 109-116 OPENDIG: CONTEXTUALIZING THE PAST FROM THE FIELD TO THE WEB ty of sites. The initial implementation of OpenDig across the three sites which make up the Madaba Plains Project (Tall Hisban, Tall Jalul and the aforementioned Tall al'Umayri) revealed that each site, using the same recording methodologies, had small variations in the interpretations and application of the forms that made it impossible to use the database at each of the three sites. This problem can be solved by using a schema-less database system such as CouchDB. CouchDB shifts away from the usual columnar databases that require a specific schema for data storage. Unlike its columnar predecessors, CouchDB is a schema-less database which calls itself a "document storage" (Anderson, Lehnardt, and Slater 2010, 4). Previously, data recorded in the field would be made by hand on a single sheet of paper, which was then digitized across several different tables. Shifting to CouchDB allows one to rethink data recording in such a way that enables the data storage to mimic the field practice; documents are created instead of tables and key-value pairs instead columns. Each document is stored in JSON (JavaScript Object Notation), which is a human-readable markup language similar to XML, but requiring significantly less overhead. Beyond schematic differences, CouchDB offers a two perks that makes it ideal for use in archaeology. First, it has a replication layer built into the database. This makes deploying the database to any number of devices a painless process. Archaeology in Jordan often takes place in remote locations with little to no access to the Internet. Often researchers will make a copy of the database that they then take to the field as they will not be able to access their primary database back home. This local copy will handle all changes to the data while used in the field and then will overwrite the database after the project returns from the field. This means that the field copy of the database becomes the primary database and all operations at the home location must cease or complex synchronization routines need to be created. Couch- 111 DB's replication layer allows for working with multiple copies of the database at one time. The built-in synchronization layer handles any conflict management and will push for "eventual consistency" (Anderson, Lehnardt, and Slater 2010, 11-20) as it handles multiple devices at any given time. The replication feature also acts as a distributed backup procedure. As long as every device using the database is synchronized on a regular basis, the database will effectively be backed up on each device, and assuming there is an Internet connection present, it can also be replicated to the cloud on a regular basis. This greatly reduces the risk of data loss through the use of cloud and local backups for replication. The application has developed over time from one single application to three integrated individual applications. In order to manage these applications efficiently, a single file defines all the fields used for recording in the field. Each of the applications then draws from this file to layout the data for display as well as entry. This way, the three applications can easily be modified from one single file, as well as adapted to other projects as each project may have different recording needs. 2.2 Part I: The Web The first OpenDig application was created solely for the web, and initially only to act as a way to publish the data. At that point, the data was still being entered into an Access database and then migrated into the Ruby on Rails web application whenever updates were made. It quickly became apparent that the web was the way forward, and it was decided to move all the data entry to the web, which was done in 2010 for the first time at the Tall al-‘Umayri excavations in Jordan. It was during that season, where all the data entry was done in the field, that the problems with dependable Internet connections became apparent, not to mention that the data entry was challenging for the supervisors who were then overloaded with both field reports and data curation. These problems inspired the © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 109-116 112 VINCENT et al creation of the first mobile OpenDig application so that data would be entered directly into the database in the field and cut out the need to transcribe the forms later. The web application also represents the most sophisticated set of tools, but this comes at the cost of a dedicated server that isn't easily deployed to the field. 2.3 Part II: The Field The original recording system at Tall al'Umayri depended on the former codirector, Larry Herr, entering all the data from the forms into a Microsoft Access database at the end of the season. Once the data were entered into the database, a processing usually lasting about two months, the database was then burned on to a DVD and distributed via postal mail to the various researchers working on the project. With OpenDig, an all digital data entry system streamlines this process by removing the need to transcribe data altogether. Figure 2 OpenDig on the iPhone. With paper-based recording systems, there was also a need to have one person verify the data on a weekly basis. This person would routinely check each notebook, making sure all the necessary data were recorded on the forms, and check for any erroneous data along the way. Just as this was a tedious task undertaken by a single individual, this process can be streamlined with OpenDig through the implementation of data validations in the application itself. In order to achieve this, an OpenDig application has been developed focused on in-field data entry using native Apple’s mobile devices such as the iPhone, iPad, or iPod. Of course, connectivity cannot be guaranteed in the field and therefore the database is built to synchronize with a database at the ‘Dig House’ after the day's excavations are completed. Data validations can be implemented to verify that correct data is being entered, while a streamlined review process can be put into place for each field supervisor to verify the necessary data has been entered for their excavation units. 2.4 Part III: The Lab Due to the connectivity problems often faced in remote areas, it is not possible to guarantee that using a remote database would work. Instead, it is necessary to use a local database with which the various field devices can synchronize their data. However, creating a server that is easy to deploy in the field isn't an easy task. Fortunately, since CouchDB comes packaged as a "one-click" install, it is easy to deploy it on any computer in the field. Furthermore, one of the most powerful components of CouchDB is the ability to host and serve applications from within the database itself. Rather than having to setup a complicated server to host the necessary application, CouchDB acts as the server, in which a lightweight data-reader and writer can be placed allowing for staff to have quick access to the excavation data while in camp. While this system doesn't have all the tools found in the main web application, it provides researchers with the ease and comfort of accessing all the excavation data from a browser. 3. AGILE ARCHAEOLOGY Agile software development (Beck et al. 2001, Martin 2003) has been key in reshaping the software development world. Acknowledging that software development is a fluid, rapidly-changing process (Martin 2003, 1-9) shifted the focus on how software is produced. Rather than trying to do an assembly-line production of software, it should instead be done in iterations, building the basics first and moving on from there. Archaeology is certainly not software © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 109-116 OPENDIG: CONTEXTUALIZING THE PAST FROM THE FIELD TO THE WEB development, but there are lessons to learn from the agile methodologies that can improve how archaeological data are recorded, shared and published. For years archaeologists have been arguing over the best way to record and share data, the format, the items that should be included, and other details relating to the sharing of archaeological data (Adam Matei, Kansa, and Rauh 2011, Atici et al. 2013, Kansa 2012, Kansa and Kansa 2013, Richards et al. 2011, Richards, Richards, and Robinson 2000, Richards 1997, Schloen 2001). Unfortunately, while these conversations are necessary and enable us to more effectively collaborate, the lack of agreement has also meant that we are still waiting for a standardized data format for archaeology, and instead we see the field in a state of fragmented data formats. However, this does not have to be a problem. The agile methodologies encourage iterations, and publishing our data quickly, even if it isn't in a standardized format, at least gets these data out and available to the research community. As more data becomes available, it should become clear how best we can share and link our data together. 4. CONNECTING DATA With disparate datasets available for research today, the question is how to link these data together to create holistic datasets for more complete research opportunities. Concepts, such as the semantic web (Berners-Lee, Hendler, and Lassila 2001) provides the ideal framework for linking these disparate datasets. Richards (Richards 2006, 977) rightly points out that using the semantic web to link archaeological data (although Richards is primarily referring to publications) still requires an agreed ontology. This holds true for archaeological datasets, but these ontologies need only be a few common fields that can be found among all archaeological datasets. Furthermore, these ontologies need only be common linking descriptions such as geographical location or chronological time period. The other data can then be returned according to a search, leaving the research- 113 er to make the final determinations regarding the relevance of these data within their greater research. As previously said, agile methodologies adapted to archaeology dictate that the primary focus is to publish our data, even if we haven't agreed on a common ontology. These ontologies can be developed and added to existing datasets as it becomes clear what these ontologies should be, particularly if we have many datasets available to see what the best ontologies might be. 4. THE UCSD CYBER-ARCHAEOLOGY ECOSYSTEM, AN EXAMPLE The Levantine and Cyber-archaeology Lab at the University of California, San Diego has focused on a geographic-centric recording system since 1999 (Levy et al. 2001). Since then, custom software has been developed to handle the archaeological data in the field, lab as well as long term research, dissemination and publication. Because of these developments, the UCSD Cyber-archaeology Lab serves as an example of both agile archaeology as well as how we might move forward towards linked archaeological data. Two additional systems, ArchField and ArchaeoSTOR, will be highlighted here and then a brief description of how these systems are being integrated. 4.1 ArchField ArchField (Smith and Levy 2012) is a system for the real-time recording and visualization of geographic data in the field. Connecting directly to total stations or GPSs, the system allows the field archaeologist to record data directly to a laptop or handheld device, visualizing and editing it in real-time, reducing the need for postprocessing in the lab after the excavations have been completed for the day. By having these data available to the researcher directly in the field, they can see what data might be missing and what data they may need to correct before leaving the field. This allows for greater accuracy for the geographical data by reducing the time be- © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 109-116 114 VINCENT et al tween acquisition and visualization. The system uses PostGIS, a SQL database with geospatial extensions allowing for the storage, indexing and retrieval of geographic data, to handle the geographic and metadata, which is then synchronized with a master database back at the lab at the end of the excavation day. chaeological data; OpenDig describing the archaeological context of these data; and ArchaeoSTOR accessing the artifacts which rely on the previous two for the context. This was carried out on three different excavation sites for the Edom Lowlands Regional Archaeology Project (ELRAP) in 2011 and 2012 (Levy et al.) 4.2 ArchaeoSTOR 5. CONCLUSION The problem of connecting disparate archaeological data collected in the field does not have to be the challenge it has been made out to be. Using agile methodologies, publishing archaeological data as soon as possible to the Web, allows researchers to begin linking archaeological data right away. Perhaps it will require more work since researchers will have to find common ontologies, however it will push the field forward as primary data is published and linked, even with additional work involved at the moment. As these data are published, common ontologies will become apparent, adding to the conversation for an archaeological data standard. Furthermore, once an archaeological data standard is agreed upon, these already published data can be updated to reflect any new standards that might be adopted in the future. In the mean time, proposed standards such as ArchaeoML (Schloen 2001) or tDAR (Plaza 2013, Kansa et al.) can be adopted as a way to bridge the gap and find common ontologies. ArchaeoSTOR (Gidding et al. 2013) grew out of the need to organize and manage artifacts, different data file formats and related data. It became apparent that the quantity of artifacts and samples being managed by the UCSD excavations was becoming increasingly difficult to manage using traditional methods. ArchaeoSTOR unified all these datasets in one place, allowing the team to quickly and easily locate and manage the artifacts, samples and analyses. By creating a management system, artifacts can be found easily, analytical data can be attached and studied quickly efficiently. 4.3 Connecting the Systems ArchField, ArchaeoSTOR and OpenDig make up the three principle systems being used in the field by UCSD's excavations. Each of the three components is independent and does not depend on the other. However, in order to conduct holistic research, all of these data are being incorporated into a single system allowing access to all three datasets seamlessly. Connecting these three distinct systems isn’t problematic if it is approached from the perspective of machine-readable data. Using Application Programming Interfaces (APIs), the three systems can be easily connected using common data. For example, similar data describing the locus context is common among all three systems. Therefore, one can simply query a single system, retrieving the data from all three systems, geographic data defining the where of the ar- 6. ACKNOWLEDGEMENTS Some of this work was supported by the National Science Foundation under IGERT Award #DGE-0966375, "Training, Research and Education in Engineering for Cultural Heritage Diagnostics", awarded to Matthew L. Vincent, who is also grateful to Thomas E. Levy for allocating his NSF travel funds to support his travel to Delphi for the Virtual Archaeology conference. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 109-116 OPENDIG: CONTEXTUALIZING THE PAST FROM THE FIELD TO THE WEB 115 REFERENCES Adam Matei, Sorin, Eric Kansa, and Nicholas Rauh. (2011) The Visible Past/Open Context Loosely Coupled Model for Digital Humanities Ubiquitous Collaboration and Publishing: Collaborating Across Print, Mobile, and Online Media. Spaces & Flows: An International Journal of Urban & Extra Urban Studies no. 1 (3), pp. 33-48 Anderson, Chris, Jan Lehnardt, and Noah Slater. (2010) CouchDB: The Definitive Guide: The Definitive Guide: O'Reilly Media. Atici, Levent, SarahWhitcher Kansa, Justin Lev-Tov, and EricC Kansa. (2013) Other People’s Data: A Demonstration of the Imperative of Publishing Primary Data. Journal of Archaeological Method and Theory no. 20 (4):663-681. doi: 10.1007/s10816012-9132-9. Beck, Kent, Mike Beedle, Arie Van Bennekum, Alistair Cockburn, Ward Cunningham, Martin Fowler, James Grenning, Jim Highsmith, Andrew Hunt, and Ron Jeffries. Manifesto for agile software development, http://agilemanifesto.org/, Accessed 5/12/2013. Berners-Lee, Tim, James Hendler, and Ora Lassila. (2001) The semantic web. Scientific american no. 284 (5):28-37. Brower, James K. (1989) Archaeological Excavation Data Management System. In Madaba Plains Project: The 1984 season at Tell el-Umeiri and vicinity and subsequent studies, edited by L.T. Geraty, LG Herr, OS LaBianca and RW Younker, 387-401. Berrien Springs, MI: Andrews University Press. Clark, Douglas R. (2011) The Madaba Plains Project : forty years of archaeological research into Jordan's past. Sheffield England ; Oakville, CT: Equinox Pub. Gidding, A., Matsui, Y., Levy, T. E., DeFanti, T., & Kuester, F. (2013). ArchaeoSTOR: A data curation system for research on the archeological frontier. Future Generation Computer Systems, 29(8), 2117-2127. Kansa, Eric. (2012) Openness and archaeology's information ecosystem." World Archaeology no. 44 (4):498-520. doi: 10.1080/00438243.2012.737575. Kansa, Eric C, and Sarah Whitcher Kansa. (2013) We All Know That a 14 Is a Sheep: Data Publication and Professionalism in Archaeological Communication. Journal of Eastern Mediterranean Archaeology and Heritage Studies no. 1 (1) : 88-97. Kansa, Eric C, Sarah Whitcher Kansa, Francis P McManamon, Keith W Kintigh, Adam Brin, and Andrea Vianello. (2010) Digital Antiquity and the Digital Archaeological Record (tDAR): Broadening Access and Ensuring Long-Term Preservation for Digital Archaeological Data. http://csanet.org/newsletter/fall10/nlf1002.html, Accessed 5/12/2013. Levy, TE, JD Anderson, M Waggoner, N Smith, A Muniz, and RB Adams. (2001) Interface: Archaeology and Technology–Digital Archaeology 2001: GIS-Based Excavation Recording in Jordan. The SAA Archaeological Record no. 1:23-29. Levy, Thomas E, Mohammad Najjar, Aaron D. Gidding, Ian W. N. Jones, Kyle A. Knabb, Kathleen Bennallack, Matthew L. Vincent, Alex Novo Lamosco, Ashley Richter, and Craig Smitheram. The 2011 Edom Lowlands Regional Archaeology Project (ELRAP): Excavations and Surveys in the Faynan Copper Ore District, Jordan. Annual of the Department of Antiquities of Jordan (in press). Martin, Robert Cecil. (2003) Agile software development: principles, patterns, and practices: Prentice Hall PTR. Plaza, David Michael. (2013) The Anasazi Origins Project Digital Archives Initiative: Transferring a Legacy Dataset to a Living Document Using tDAR, http://core.tdar.org/document/391353, Accessed 5/12/13. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 109-116 116 VINCENT et al Richards, Julian. (2006) Archaeology, e-publication and the semantic web. Archaeology no. 80 (310):970-979. Richards, Julian C, Julian Richards, and Damian Robinson. (2000) Digital archives from excavation and fieldwork: a guide to good practice: Oxbow Books Ltd. Richards, Julian D. (1997) Preservation and re-use of digital data: the role of the Archaeology Data Service. Antiquity no. 71 (274):1057-1059. Richards, Julian, Stuart Jeffrey, Stewart Waller, Fabio Ciravegna, Sam Chapman, and Ziqi Zhang (2011) The Archaeology Data Service and the Archaeotools project: faceted classification and natural language processing. Archaeology 2.0: New Approaches to Communication and Collaboration : 31-56. Schloen, J.D. (2001) Archaeological data models and web publication using XML. Computers and the Humanities no. 35 (2):123-152. Smith, Neil G, and Thomas E Levy. (2012) Real-time 3D archaeological field recording: ArchField, an open-source GIS system pioneered in southern Jordan. Antiquity no. 86 (331). http://antiquity.ac.uk/projgall/smith331/, Accessed 5/12/2013 © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 109-116 Mediterranean Archaeology and Archaeometry, Vol. 14, No 4, pp. 117-123 Copyright © 2014 MAA Printed in Greece. All rights reserved. EDUCATIONAL ENHANCING OF VIRTUAL EXPOSITIONS. TOWARDS VISITOR-CENTERED STORYTELLING DIGITAL MUSEOLOGY Ioannis Kanellos, Simona Antin, Orestis Dimou and Maria-Anastasia Kanellos Computer Science Department, Télécom Bretagne, CS 83818, 29238 Brest Cedex 3, France Received: 09/12/2013 Accepted: 22/06/2014 Corresponding author: Ioannis Kanellos (ioannis.kanellos@telecom-bretagne.eu) ABSTRACT We present here some basic ideas and functions of a system developed for cultural, scientific and technical mediations of a good quality but still affordable adapted storytelling. The adaptation concerns mainly the topic variety and the knowledge deepness. In the first part we discuss the origins of the storytelling approach, its interest in mediation tasks; we also handle the problem of transformation of some narrative functions into operational categories of analysis by means of adequate metadata. In the second part, we give some elements of the knowledge organization underlying such a system and precise the notions of learning profile and of point of view, for a case study concerning a videobased presentation of a Renaissance painting; in our case, we distinguish three profiles and nine points of view. We then give some examples illustrating the capacity of the system to build abundant adapted stories about the chosen painting; we also argue in favour of the local-ontology construction, which supports the whole system, both at the backand the front-office levels. We finally conclude with some educational enhancements we currently develop (storytelling generalization containing complementary mediation modes, as Problem Resolution and Serious Games, integration of stereoscopic 3D videos, intelligent track of an end-user during her/his reading process as well as customized contextual assistance to her/him reading goal). KEYWORDS: digital museology, storytelling, cultural and educational mediation, adaptive systems, variable deepness knowledge representation, user profile, points of view, ontologies, reading and interpretation strategy EDUCATIONAL ENHANCING OF VIRTUAL EXPOSITIONS 1. INTRODUCTION Since at least two decades, storytelling stimulates, and this rather intensively, countless innovating ideas in the broad field of cultural, scientific or technical mediation. Education and cultural/scientific heritage are equally addressed. There are of course efficient and final causes for such a state of affairs (Aristotle, Physics, II, 3, 194b 23-195b 21, Metaphysics, IV 2, 1013a 23 sq.). As semiotic consumers, can we say, we definitely are beings of narration. Presumably, because our modern thought and, generally, our modern culture, have never been deserted by their mythic origins (for instance: Vernant 1990, 2006; Lévi-Strauss, 1995, between others.). In various domains, where mediation becomes a central affair (Lamizet 2000, Guillaume-Hofnung 2011), education, museology and generally, cultural and scientific heritage, we progressed (or retro-progressed), moving from reason to myth. We discovered both the myths of our reason and the reasons of our myths. Storytelling stands nowadays more than a concept: in fact, it became a genuine productive tool facilitating knowledge acquisition, fostering interpretive and problem solving abilities, allowing rich sharing of cultural experiences... To put it briefly: a powerful means supporting, sustaining and refining our need for comprehension. In all models and applications, storytelling appears as a major vector of knowledge and culture transmissions. Myriads of projects give statistical evidence of a new epistemological conscience, which corresponds to a well-identified shift in storytelling cultural economies (Bedford 2001, Leinhardt et al. 2002, Puhol et al. 2013; Filippini-Fantoni 2004; cf. also the European project CHESS). Certainly, as a practice, storytelling goes farther than conventional textual narration; it splits up quickly into numerous species engaging enaction, interaction and diverse participative activities, both individual and social. It appears nowadays as an all-purpose rationalizing method, certainly of an unbalanced and nomadic nature, but still 118 operational, where fissures of significance, incongruities, interruptions, multi-level associations, plot reorientations, etc., are still possible, plausible and even essential in serving understanding, transmission and sharing of meaning. One expects various added values in exploit it in research and development of systems aiming at facilitating cultural, scientific and technical mediation. Natural and motivational, definitely efficient, storytelling protocols are nevertheless laborious. For a large-scale development, they become even unaffordable, as far as they have to be adapted to various reception modes, to unequal cognitive skills, to different horizons of expectations and, generally, to diversified noetic requirements. Astonishingly, even in the heart of a profound economical crisis, the cost of such developments seems less interesting for academic research. But the storytelling problem turns, anyway, and critically, into a problem of adaptive low-cost industrial design. It better emerges as an industrious, systematic but realistic endeavour: is it possible to build refined storytelling alternatives that fit to different reception capabilities expressed by different semiotic consumers? Behind of such a plain formulation, one immediately understands the productive challenge; it reformulates the need for a respective, wide-ranging, popular, and perhaps self-governing or even autonomous access to knowledge and, generally, to culture (Braz 2011, Bourdieu 1979, Caune 2006). Based on some previous experiences concerning profile-adapted thematic virtual museums (Kanellos 2009, or The Annunciation virtual museum, for instance), our contribution concerns, precisely, the outline of a project we currently develop concerning a system that allows cost-effective conceptions and uses for differentiated narration scenarios. The intent is to satisfy various demands of storytelling. More generally, the system aims at furnishing a flexible platform able to set up virtual presen- © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 117-123 EDUCATIONAL ENHANCING OF VIRTUAL EXPOSITIONS tations of an item (or a theme) or of a collection of items (or themes) for different visitor profiles, for different learning goals, and for different educational levels; in other words, for different reading and interpretation strategies. Here applied to an art history case study, the system supports actually any virtual exposition logic. Such logic is significantly augmented with adapted “intelligent”, assistance in reading and interpretation, allowing alternative comprehensions of the exposed material. Moreover, the system offers various observation enhancements that may lead to more refined meaning appropriations. In a certain sense, through adapted storytelling, such a system tries to grasp the progress in understanding as a progress of accommodated narration: the object or theme is presented under a tailored form, fitting to defined reception capacities and expectations. 2. KNOWLEDGE ORGANIZATION FOR MULTIPLE STORYTELLING: FUNDAMENTALS In a theoretical lineage traceable to narratology and, more specifically, to the study of (Propp 09), the system we present substitutes to the notion of narrative function the more operational one (for a presentation) of point of view. A point of view defines a particular approach on the subject, using categories of thought that are broadly irreducible to the remaining points of view (for instance, aesthetics, contextual support, comparative approach, hermeneutics, etc.). A narrative unit is seen as a concatenation of partial “stories” (all of them worked up under the same linguistic register), addressing some points of view (topics) concerning the subject. In our case study, dealing with a painting (The Flagellation of Christ, by Piero della Francesca, ~1455, Galleria Nazionale delle Marche, in Urbino, Italy), the elected points of view are basically seven: • Description • Author 119 • Context • Documentation • Aesthetics • Technical Information • Interpretation For a complete and somehow natural presentation, we added to this list an Introduction and a Conclusion; they also can be seeing as points of view, rising the total number of points of view to nine. The number and the nature of the points of view are not mandatory: different storytelling conceptions usually necessitate different points of view. The back-office of the system we develop (indexing module) offers a convenient environment in order to define an arbitrary number of points of view, customizing various conceptions on exposition. Each point of view is divided into consecutive deepness levels, which correspond to reception refinements; they address equinumerous user profiles. Clearly, one can imagine an unlimited list of profiles. The notion of profile seems necessary, as far as the requirements for comprehension, and thus, for analysis, may differ a lot from a learner to another (see, for instance, (Arasse 2005)). We limited our study to three profiles, somehow emblematic in the cognitive literature (corresponding to stages in knowledge acquisition): 1. amateur (that implements general public demands), 2. confirmed (addressing needs of an academic student) and 3. expert (satisfying scholar requirements). Profiles are indexed to local ontologies and design levels of apprehension ability. For instance, contrary to what is laid down in the CHESS project (cf. above), where profiles are rather informal and correspond to sociological and/or educational categories of museum visitors, in our case, profiles, are defined formally as guides to planning by-level learning requirements. All contents we deal with are video sequences, lasting from some seconds to sev- © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 117-123 EDUCATIONAL ENHANCING OF VIRTUAL EXPOSITIONS eral minutes; each of them furnishes matter to particular aspects of the painting. (The total duration of all these sequences is about one hour.) Clearly, the nature of these contents is unintentional: any multimedia resource may be used instead. In the light of these principles of analysis, the entire storytelling material is partitioned off, giving rise to a two-dimensional matrix, where levels of refinement (profiles) cross points of view (topics with adapted narrativity). Any exposition (of the concerned object or theme) is actually a choice function over this matrix. 120 all points of view and all profiles are upstream defined; and that the whole base of video resources is available. Figure 2 gives a snapshot of this end-user interface. Figure 2: Left: choice of the video resources through the Profile/Point-of-view matrix. Right: video player (above) and storytelling editor (below). Figure 1: Points of view are topics on the subject. Levels stand for profiles. As far as a particular exposition invokes at least one point of view and one profile, in our rather limited case study (with 9 points of view and three deepness levels), and excluding possible narrative permutations (i.e. narrations made up from the same set of video sequences but in a different order), the system can generate more than 250000 by-topic and by-profile cus9 tomized “stories” (4 -1 more precisely) for this particular painting of Piero della Francesca. 3. INTERACTION FEATURES For a better comprehension of the use, we give in this section some elements concerning the interaction with the interface. Our illustration takes aim at a front-office user (typically, a professor composing a course for a class, a curator preparing an exposition, or even a student or a museum visitor wishing to learn more about a subject). Clearly, in this case, we suppose that Aiming at setting up her/his exposition or presentation, the user picks out the story components from the matrix, by adjusting their nature and level to her/his target auditory (left). Narrative coherence is resolved upstream, when conceiving the resources. The most natural and, perhaps, low-cost way is to design profiles inclusively, i.e. so that the L+1 includes the L profile and is included in the L+2 profile. A textual summary offers quick information on each resource that can also extensively be viewed through the player (right). Selected resources can be edited at any moment through the storytelling editor, below the player (giving, thus, a global “plot” organization). Complementary metadata furnish information about the point of view used, the profile concerned, the duration (of the atomic video resource or of the global exposition/presentation storytelling already built), etc. The user may also “augment” this story by adding her/his own comments. An annotation editor allows such possibility (Figure 3). © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 117-123 EDUCATIONAL ENHANCING OF VIRTUAL EXPOSITIONS 121 each one corresponding to a specific point of view. Figure 4 gives a contracted version of the Aesthetics ontology (the number of concepts of this point of view is about 200). Figure 3: Annotations may be added at any narrative moment of the story under construction. Annotations may be viewed at the same time with the video sequences. They supply comments and, generally, complementary information to the global story already built. 3.1 Visualization. Sharing and social contributions. Figure 4: Part of the local ontology corresponding to Aesthetics point of view. Once the global scenario is completed, it may, of course, be viewed and/or be saved as such. Authorized persons (students, visitors...) can then view the “storified” version that presents the object/theme to their level and goal. At any time, it is possible for the designer of the presentation (teacher, curator, etc.) to come back and (re)edit the story already built. At any time, it is also possible (when modification rights are provided) to take an existing story (made up by someone else) and slightly transform it in order to adapt it to some new reception goal. On the other hand, the system offers possibilities for sharing the storytelling experience both in building and in viewing. This last leads us to give some hints about the back-office of the system. In our case, the total number of concepts for all points of view is about 2000. The hierarchical ontology mode is based on a partial order relation for optimization reasons (large-scale thorough performance). The tree structure of such ontologies gives ground for low-cost indexing processes based on similarity features of the items concerned by the presentation or the exposition (in our case, paintings). The back-office user is a key-user insofar as she/he determines the operational categories that will support, downstream, storytelling creations. Besides indexing facilities, the system allows to define arbitrary (in nature and number) points of view and profiles. Ontologies need, of course, to correspond to these definitions that fix the conceptual framework and the targeted public of the foreseen customized presentations or expositions. 3.2 The knowledge organization The knowledge base supporting the system is an aggregation of local ontologies, © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 117-123 EDUCATIONAL ENHANCING OF VIRTUAL EXPOSITIONS 122 4. FURTHER WORK. CONCLUSION. The project we briefly presented is actually a part of a bigger project that aspires to develop a “Builder” for general-purpose virtual expositions. So far, we presented the narrative part of the system that roughly corresponds to the storytelling initiative in presentations. But, as we had already the opportunity to mention it, storytelling may also embrace alternative modes of knowledge transmission, essentially learning by projects and other general active pedagogy techniques, gaming, etc. The two-dimensional matrix we saw in section 2 is then complemented by a third dimension, where one defines alternative pedagogical modes. Any narration mode may invoke, therefore, corresponding activities. A second direction of research and development concerns the integration of stereoscopic 3D videos. Indeed, more and more, both, the e-class and the e-museum, open to new stereoscopic technologies that reinforce motivational traits through enactive features. We presently work on some extensions of the system we presented, taking into account such possibilities. Figure 5: A storytelling approach composing with various mediation modes (especially Problem Resolution and Ludo-Educative Activities). Finally, a third direction concerns the elaboration of a semantic model of the student/visitor in order to track her/his activities in using the platform in order to render to her/him contextual relevant assistance for profile-adapted learning. The ontologies we presented are essentially thought out with this goal in mind. In some few and rather resuming terms, the storytelling affair is, in effect, a powerful paradigm to conceive and develop solutions that renew the transmission of knowledge and culture in our digital area. The system we try to set up answers to a generalized demand merging cultural, scientific and technical mediation techniques in an integrated storytelling based platform. ACKNOWLEDGEMENTS The project we presented is part of a larger project, dealing with narrative approaches of teaching by means of 3D stereoscopic video sequences (Education 3D); it is financed by the French Ministry of Education. REFERENCES Annunciation Virtual Museum: www.annunciation.gr Arasse, D. (2005): On n’y voit rien. Descriptions. Paris, Denoël. Bedford, L. (2001): Storytelling: the real work of museums. Curator 44(1): 27-34. Bourdieu, P. (1979): La distinction : critique sociale du jugement. Paris: Éditions de Minuit. Braz, A. (2011): Bourdieu et la démocratisation de l’éducation. Paris, Presses Universitaires de France. Caune, J. (2006): Culture et communication : convergences théoriques et lieux de médiation. Grenoble, Presses Universitaires de Grenoble. Chess project: http://www.chessexperience.eu © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 117-123 EDUCATIONAL ENHANCING OF VIRTUAL EXPOSITIONS 123 Filippini-Fantoni, S. (2004): Personalization through IT in Museums. Does it really work? The case of the marble museum website. In ICHIM03: Cultural institutions and digital technology. Paris, École du Louvre. Guillaume-Hofnung, M. (2011): La médiation. Paris, Presses Universitaires de France. Hooper-Greenhill, E. (2009). Museums and Education: Purpose, Pedagogy, Performance. Kindle Edition. Kanellos, I. (.2009): Les musées virtuels et la question de la lecture : pour une muséologie numérique centrée sur le visiteur. Revue des Interactions Humaines Médiatisées 10 (2), pp. 3-33. Lamizet, B. (2000): La médiation culturelle. Paris, L’Harmattan. Leinhardt, G., K. Crowley, K., Knutson, K. (2002): Learning Conversations in Museums. Routledge. Lévi-Strauss, C. (1995): Myth and meaning. Cracking the code of culture, New York, Schocken Books. Propp, V. (2009): Morphology of the folktale. Austin, University of Texas Press (first edition in the USA, 1968). Pujol, L., Katifori, A., Vayanou, M., Roussou, M., Karvounis, M., Kyriakidi, M., Eleftheratou, S., & Ioannidis, Y. (2013): From personalization to adaptivity. Creating immersive visits through interactive digital storytelling at the Acropolis Museum. In J. A. Botía & D. Charitos (Eds.), Museums as intelligent environments workshop, Proceedings of the 9th International Conference on Intelligent Environments, pp. 541–554. Athens, Greece, IOS Press. Vernant, J-P. (1990): Myth and society in Ancient Greece. New York, Zone Books (first edition in French, 1974). Vernant, J-P. (2006): Myth and thought among the Greeks. New York, Zone Books (first edition in French, 1965). © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 117-123 Mediterranean Archaeology and Archaeometry, Vol. 14, No 4, pp. 125-133 Copyright © 2014 MAA Printed in Greece. All rights reserved. THE ART OF IMPLEMENTING SFM FOR RECONSTRUCTION OF ARCHAEOLOGICAL SITES IN GREECE: PRELIMINARY APPLICATIONS OF CYBER-ARCHAEOLOGICAL RECORDING AT CORINTH Thomas E. Levy1, Matthew L. Vincent1, Matthew Howland1, Falko Kuester2, and Neil G. Smith3 1Department of Anthropology and Qualcomm Institute, University of California, San Diego, USA 2Department of Structural Engineering and Qualcomm Institute, University of California, San Diego, USA 3Visual Computing Center, King Abdullah University of Science and Technology, Thuwal, Saudi Arabia Received: 18/12/2013 Accepted: 25/06/2014 Corresponding author: Thomas E. Levy (tlevy@ucsd.edu) ABSTRACT Cyber-archaeology represents the marriage of archaeology, computer science, engineering, and the natural sciences with the aim of taking advantage of constantly evolving technologies for digital data capture, curation, analyses and dissemination. Digital data collection tools are perhaps the most rapidly changing arenas of development in cyber-archaeology and are becoming affordable tools for every archaeologist. In this paper, we examine two users’ approaches to produce point cloud models of archaeological sites using structure from motion (SfM) photography. The experiment took place at the Fountain of Peirene in ancient Corinth, Greece. Their implementation of the technology and their results are compared to highlight the very important role the photo-shooting session can play in the final outcome of the SfM reconstruction. We correlate the users’ approaches to the applied algorithms’ robust features and known limitations to provide a technical explanation of how archaeologists can significantly improve their success in SfM. As new algorithms and software emerge making SfM a common tool in archaeological documentation the methodology presented in this paper will enable archaeologists to meet the high demand for digital documentation on a global scale. KEYWORDS: Cyber-archaeology, digital data collection, Structure from Motion, SfM, Corinth 126 THOMAS LEVY et al 1. INTRODUCTION The past decade has seen an exponential growth in interest by researchers in the application of digital technologies to archaeological research. The reason for this has to do with the vast quantities of data collected during the archaeological process; the relatively low cost of digital tools such as surveying equipment, computers, portable storage units, cameras, and more; and the need to find ever more efficient ways of collecting, storing, analyzing and sharing those data. Cyber-archaeology provides a solution by developing an integrated system for data capture, curation, analyses and dissemination using traditional print-based systems, 3D visualization and the internet (Figure 1; Levy 2013). The University of California, San Diego Levantine and CyberArchaeology Laboratory has contributed to the development of cyber-archaeology since its Edom Lowlands Regional Archaeology Project (ELRAP) in Jordan ‘went digital’ in 1998 (Levy, et al. 2001) developing an integrated geo-spatial collection and curation for field archaeology (Levy and Smith 2007). The Greece Cyber-Archaeology Collaboratory Project was initiated in September 2013 to compare, contrast and improve digital methodologies used in projects with longterm, sometimes full-time, research endeavors with those developed by ELRAP that is characterized by two-month long excavations over one to three seasons. The American School of Classical Studies in Athens (ASCSA) is an ideal partner to organize this collaboration. Founded in 1881 by a consortium of American Universities, the ASCSA began excavations at Ancient Corinth in 1896, which are one of the oldest continuing excavations in the world. Excavations at Corinth have revealed a vast Roman metropolis that for five centuries was one of the most important cities in the ancient Mediterranean world. As our field season in Greece was only five days, we deployed two data capture technologies – Structure from Motion for rapid 3D documentation of ancient monuments and CAVEcam stereo photography that are described below. Here we report only on the SfM results from Corinth. Figure 1 Model of Cyber-Archaeology system. 2. EXPEDIENT 3D DATA CAPTURE – STRUCTURE FOR MOTION (SFM) AND BEST PRACTICES We used Structure from Motion (SfM) to create rapid 3D digital models of architectural complexes on the Greek mainland using a variety of DSLR cameras (Nikon D80, Canon 50D, and Canon 30D). SfM reconstructed models provide a situated 3D context for all the artifacts, architecture and loci we record using ArchField (Smith and Levy 2012) or other GIS-based recording systems. ‘Structure from Motion’ refers to the method of extracting a 3D structure from many overlapping digital images. Rather than standing in a fixed position and capturing 3D data, the method uses a change in camera position for each image to find the distance (motion) between them and at the same time triangulate the 3D positions of pixels matched in overlapping images. The more motion and movement around the site, the more complete the 3D model becomes. SfM is composed of several computer vision algorithms. 1 ScaleThere are many variants of the computer vision algorithms used in the SfM pipeline. Proprietary software such as AgiSoft Photoscan ((http://www.agisoft.ru/) and Pix4D (http://pix4d.com/) use their own algorithms. 1 © University of the Aegean, 2014, Mediterranean Archaelogy & Archaeometry, 14, 4 (2014) 125-133 THE GREECE CYBER-ARCHAEOLOGY COLLABORATORY PROJECT (GCACP) invariant feature transform is used to automatically detect unique features in the images (Lowe 2004). SIFT creates descriptors for each feature that enables it to locate the same feature in other images even when there is a change in scale, rotation, position, or lighting. Bundle Adjustment uses these matched SIFT features and an estimate on the camera focal length to solve for the camera position, rotation, radial distortion, and actual focal length for each image using least squares approximation (see Snavely et al. 2006). As it estimates camera positions for each image it is estimating the 3D position of the matched SIFT features. These features become a sparse set of 3D points representing the general structure of the captured scene. As a final stage a MultiView stereo algorithm like PMVS (Furukawa and Ponce 2007) is used to generate a dense collection of 3d points using the now calibrated position of the images and the sparse 3D points. 2 The collection of matched pixels and their calculated 3D positions become a cloud of millions of 3D points, called a point cloud. From a distance the point cloud appears as a solid model similar to 3D models seen in CAD programs or video games, but as you zoom in it becomes clear it is actually a collection of millions of points. With SfM, between 5-20 million 3D points are captured, enabling the recreation of excavation surfaces and architecture digitally. Although the resolution is much lower than a laser scan, it is much faster, easier to perform, and vastly more accurate than hand illustrated plans. In order to meet the demands of archaeological In this paper we used VisualSfM, an open source software that uses siftGPU (Wu 2007) and MPBA (Wu et al. 2011) for accelerated GPU processing. 2 There are two main approaches to MVS: 1. Patched based PMVS (Furukawa and Ponce 2007), CMPMVS (Jancosek et al. 2011); 2. SemiGlobal Matching (SGM) (Hirschmüller 2008) used by SURE (Rothermel et al. 2012). 127 documentation, we have been working at UC San Diego and KAUST to push SfM’s capabilities to its limits (Levy, et al. 2012). 2.1 Ancient Corinth The site is located in the northeast corner of the Peloponnese at the head of the Gulf of Corinth and was referred to in antiquity as one of the fetters of Greece, guarding as it did the narrow land bridge that connects the Peloponnese with the Greek mainland, and providing access to both the Gulf of Corinth to the north and the Saronic Gulf to the east. This strategic position was one of the keys to its prosperity, especially as a Roman city. Excavations by the American School of Classical Studies at Athens have continued for over a century with little interruption until today. The primary focus of excavations has been on the area of the Roman Forum, located within the west side of the village and south of the hill surmounted by the mid-sixth c. B.C.E. Temple of Apollo. This dominating monument has been one of the only features of the site visible since antiquity. For our SfM experiments, we focused on the Fountain of Peirene, an impressive monument described by the Greek historian Pausanias, and according to Greek myth, was a favorite watering hole for Pegasus, the winged horse that was the offspring of Poseidon, the gods of water and earthquakes, and Medusa, the Gorgon female creature. The Fountain is east of the Agora at Corinth and one of the most important fountains of ancient Greece. It represents a challenge for any 3D imaging project because it contains remains from many construction periods with different features ranging from a façade of natural rock to free-standing marble columns from the Byzantine period (Robinson 2011) so it is difficult to parse out these metadata in a scan. Although SfM is being rapidly adopted by archaeologists and cultural historians due to the development of user-friendly software such as Agisoft Photoscan (http://www.agisoft.ru/), the quality and © University of the Aegean, 2014, Mediterranean Archaelogy & Archaeometry, 14, 4 (2014) 125-133 128 THOMAS LEVY et al accuracy of the results is still predominately dependent on the quality of the images and the capture methodology employed. As an experiment we captured the Fountain of Peirene with two different cameras (Nikon D90 and Canon EOS 30D) and two users. In this way, we can compare the two final reconstructions, pinpoint how different users approach the application of the method and determine what practices lead to the best results in light of the known unresolved computer vision problems in SfM. 3. RESULTS The results of both users’ acquisition of the fountain of Peirene using SfM appear at first glance to have captured a significant portion of the fountain’s architecture (figs 1-2). However, under close examination it becomes clear that User A’s reconstruction was much more complete but still had several large gaps and point cloud coloring issues. Below we detail the main differences between the two users and explain how the user’s capturing methodology could have achieved better results. When the users cameras and chosen settings are compared it would appear that User B would acquire a much higher quality scan. In some respects this turned out to be the case. User B had a higher focal length and higher resolutions sensor compared to User A (Table 1). We can directly compare ground sampling density of the two users’ camera setup by calculating their field of view taking into consideration each camera’s crop factor and sensor resolution. At five meters User B’s ground sampling density is 8.5 pixels/cm which is ca. 1.5x the density in pixels of User A’s camera setup. However, the wider field of view allows User A to capture more area in each picture. In figure 2, where both user’s photographed from the same position it is clear that User A was able to capture more area. Table 1: Comparison Chart of Cameras and SfM Results Category User A User B Camera Canon EOS 30D Nikon D90 Focal Length 18mm (28.8mm)* 24mm (36mm)* Resolution 3522x2348 4310x2868 Pixel Density at 5m 5.6px/cm 31.6px/cm² 8.5px/cm 72.25px/cm² Camera Orientation Portrait Portrait Camera Setting Aperture Priority f/8 Aperture Priority f/4 Pictures Taken 478 548 Pictures Matched 473 349 Picture Efficiency 98.95% 63.69% Sparse Point Features 233,003 203,139 11,216,589 12,821,729 48 63 10,955,837 12,289,853 260,752 531,876 95% 75% Dense Point Cloud (PMVS) Dense to Sparse Points1 Dense Points after Cleaning Excess Points Removed Qualitative Completeness2 1Calculates how many dense points could be extracted from images given found sparse points. A product of MP of camera. 2Area Captured/Total area of structure. *Equivalent Full frame focal length This translates into more features that can be automatically detected for matching across images but at a slightly sparser density. User A’s photograph resulted in 39,669 SIFT features, while User B’s photograph had 46,780 SIFT features. User B had a denser count of features within a smaller field of view. User A’s camera setup is best geared towards capturing more area which will help during matching and the final bundle adjustment of the © University of the Aegean, 2014, Mediterranean Archaelogy & Archaeometry, 14, 4 (2014) 125-133 THE GREECE CYBER-ARCHAEOLOGY COLLABORATORY PROJECT (GCACP) reconstruction. In contrast User B’s camera setup will have a higher density of extracted SIFT features per image, but User B will have less area to match between images unless they take more pictures than User A with more overlap. 129 denser in this area for User B. In figure 4, a close-up of the steps, shows that User B’s point cloud has a greater density. However, despite the greater point cloud density the coverage and qualitative results of User B’s to User A’s scan are much lower in part due to the smaller FOV and other acquisition mistakes discussed further below. Figure 2 First images with starting position for both Users. This figure shows a good comparison of differences between lenses and camera sensors. Figure 4 A close-up of the steps and white marble in background shows that User B (Bottom) acquired a denser point cloud due to a higher megapixel camera and higher focal length. Figure 3 User A (Top), note more complete reconstruction of surfaces. User B (Bottom), more lost data but areas appear sharper. An analysis of the final reconstructions for both users highlights the trade-off between field of view and ground sampling density (GSD). For example, User B’s reconstruction had a slightly higher point cloud density (1,334,016 more points) due to the higher image GSD (see table 1). In figure 3, where User B captured the same scene in almost the exact same positions the 3D point cloud appears sharper with brighter and more defined colors because the point cloud is much When closely examining the individual input images and resulting point clouds it becomes clear that neither user compensated for light changes and the effects of their cameras’ built in white balance meter. First, many of the images where depth-of-field played a role were not pixel sharp, the combination of aperture and shutter speed resulted in a shallow focus. Second, both users periodically had blurred pictures due to low shutter speeds as they were capturing and moving at the same time. Lighting problems are most apparent when they faced the dark clouds illuminated by the sun. The result for both users was dark under exposed images (figures 5-6). © University of the Aegean, 2014, Mediterranean Archaelogy & Archaeometry, 14, 4 (2014) 125-133 130 THOMAS LEVY et al Figure 5 Under exposed areas occur for both Users. User B’s (Right) automatic settings led to much more under exposure. User A: 18mm, 1/60, f/8, ISO100. User B: 24mm, 1/400, f/4, ISO-200. Figure 6 The effects of under exposed images can be seen in these figures, where the pillars and back wall appear to have a false shadow. Note User B’s reconstruction of this area is sparser and missing sections of the recessed courtyard. This again directly affects the amount of SIFT features that can be detected and matched. SIFT and other related algorithms are typically quite robust to minor light change but extreme illumination changes or lack of sufficient light are quite disruptive (c.f. Lowe 2004). Dark shadows and under exposed images will have significantly less features and in turn fewer successful matches. Examining the amount of sift features extracted for the underexposed images in Figure 5, User A’s photograph had 30,988 SIFT features, while User B’s had 11,861 SIFT features. The significantly reduced count of SIFT features compared to User B’s average (ca. >45,000) is directly related to the underexposure of the image. Although these images were still included in the complete reconstruction, the affect of the lighting can be seen especially in some of the pillars, which in the case of User B were almost black (where no points were extracted). In general, lighting for both users was highly variable; this is likely due to over-reliance on Matrix Metering (Nikon) or Evaluative Metering (Canon) that processes the entire scene’s lighting. Spot metering, relying on a single, or small, area helps prioritize the exposure to the user’s desired target. A qualitative comparison between the two users indicates that although User B had a significant advantage in the selection of camera, a number of mistakes were made leading to User A being much more successful in conducting a thorough capture of the fountain (figures 7-8). Although both users maintained a fixed focal length for the majority of capture and User B roughly followed User A, their results differed significantly. Since SfM algorithms use a positional change in camera position for each image to find the distance (motion) between them it is critical to move and not stay in one place, maintain significant overlap and not make large rotational changes all at once. When capturing complex archaeological sites a systematic approach should be planned before hand to ensure the capture session achieves full coverage and mistakes in maintaining image overlap are not made when transitioning from one area to another. User B failed to achieve enough overlap to fully reconstruct the entire fountain due to a lack of planning and not fully understanding the limits of the SfM algorithms. First, the southern section had few overlapping images and resulted in it being processed as a different model (Figure 9). Second, User B often stood in the same position and rotated the camera (panoramas): these were found during the bundle adjustment but did not contribute © University of the Aegean, 2014, Mediterranean Archaelogy & Archaeometry, 14, 4 (2014) 125-133 THE GREECE CYBER-ARCHAEOLOGY COLLABORATORY PROJECT (GCACP) 131 significantly to key feature matches since they had little parallax. Third, User B’s images overall were much darker resulting in a poorer quality reconstruction and possibly led to the loss of key tie points. User B’s reconstruction had more points in the model due to the higher GSD, but qualitatively it is difficult to determine where this higher density paid off. Certain areas are more detailed in User B’s reconstruction, but overall it had many more holes and areas too underexposed to be clearly seen in the final point cloud. User B moved parallel to each side of the square courtyard (figure 9). Rarely were shots taken at oblique angles, another possible cause of poorer and more occluded capture. Figure 8 Dense Reconstruction of Fountain of Peirene. Note many more sparse areas in User B (Bottom) reconstruction and missing right fountain (Processed by PMVS). Figure 7 Sparse Reconstruction and calibrated camera positions for Fountain of Peirene. User A (Top), and User B (Right), processed in VisualSfM. In contrast, user A appears to be more experienced with SfM acquisition and had a specific plan of how they would sufficiently capture the entire site. Matching of image features is robust to 30 degrees in any direction (Lowe 2004). The user had very thorough coverage paying attention to not turn sharp angles (>30 degrees) and insuring significant overlap between each image (figure 9). At corners extra shots were taken and the user appears to have turned 90 degrees to shoot down the path they came and also about faced to get close-ups of the corners to better tie them in. User A spent special attention to difficult areas and took many detailed shots of areas with high occlusion or windows/pits. In summary, even though User A took fewer pictures with a lower GSD the method of documentation resulted in a much better capture than that of User B. Both users’ results could have been even better if they paid greater attention to lighting, shutter speed, and aperture. Finally, both users failed to adequately capture the floors to reconstruct them properly using Patch-based Multi-view Stereo (PMVS) software and would be an © University of the Aegean, 2014, Mediterranean Archaelogy & Archaeometry, 14, 4 (2014) 125-133 132 THOMAS LEVY et al ideal candidate for an improved Multiview stereo algorithm called CMPMVS (Jancosek et al. 2011). The ability to point the camera down into the inset courtyard enabled this area to be captured but the other areas horizontal to the camera were sparsely documented (see Figs. 6-8). Especially in the most important area (the fountains), neither user angled the camera down to the floor to capture this tricky area. Figure 9 Close-up of the sparse reconstructions and calibrated camera positions for Fountain of Peirene, User A (Top Left); User B (Top Right) (Processed in VisualSfM (Wu et al. 2011)). The results show that opportunistic capture is not as much a threat to Cultural Heritage as one might think. Rouge SfM photographers cannot be compared to a trained surveyor with ample time to wait for the best lighting conditions, plan a detailed approach for full coverage of the site, and conduct follow-up visits to address mistakes in their photo shooting session. 4. CONCLUSION SfM provides an ideal 3D solution to rapidly recording monuments uncovered by long-term excavation projects such as the ASCSA project at Corinth. Implementing proper capture of cultural heritage sites using SfM is a learned art. It has been argued here that the proper setup and methods applied during an SfM photoshooting session play a very critical role in the final outcome. Archaeologists seeking to apply this method must take into consideration the limitations of the algorithms used in SfM and apply a systematic approach to full site coverage. A developed methodology for applying SfM in the field plays as much if not more influence on the final reconstruction than the specific SfM program used. In the near future, we will work toward using SfM as a ‘digital scaffold for embedding many years of metadata collected by the generations of excavators who have worked at this remarkable site. While it took less than two hours to capture the Fountain of Peirene using SfM technology, the ease of capturing such data does not make it right to carry out such work without permission of the authorities, whether in Greece or any other country. Ease of data capture raise hard ethical issues about who owns the cultural heritage digital datasets that are becoming so easy to collect. Ultimately, we believe the same ethical standards apply to those professionals wishing to capture digital data at cultural heritage sites as to those researchers that wish to study any aspect of the patrimony of a country. Consequently, this will always begin with obtaining a permit from the authorities of the country where such work is to be carried out. 5. ACKNOWLEDGEMENTS We are grateful to James C. Wright, Director, of the American School of Classical Studies in Athens for securing permissions and travel to Corinth to carry –out the SfM © University of the Aegean, 2014, Mediterranean Archaelogy & Archaeometry, 14, 4 (2014) 125-133 THE GREECE CYBER-ARCHAEOLOGY COLLABORATORY PROJECT (GCACP) 133 experiment described here. Thanks to Guy Saunders, Director of the American School of Classical Studies’ Excavations at Corinth for facilitating our work at the site. Special thanks to the Institute for Aegean Prehistory (INSTAP) for a travel grant awarded to Thomas E. Levy that facilitated group travel to Greece for this project. Matthew L. Vincent was supported by the National Science Foundation under IGERT Award #DGE0966375, "Training, Research and Education in Engineering for Cultural Heritage Diagnostics." REFERENCES Furukawa, Y., and J. Ponce. (2007) Accurate, Dense, and Robust Multi-View Stereopsis. IEEE Computer Society Conference on Computer Vision and Pattern Recognition, July 2007. http://grail.cs.washington.edu/software/pmvs/bibtex.txt Hirschmüller, H. (2008) Stereo processing by semiglobal matching and mutual information. IEEE Transactions on Pattern Analysis and Machine Intelligence 30, pp. 328–341. Jancosek, M., T. Pajdla. (2011) Multi-View Reconstruction Preserving Weakly-Supported Surfaces. CVPR 2011 - IEEE Conference on Computer Vision and Pattern Recognition 2011. Levy, T. E., Anderson, J.D., Waggoner, M., Smith, N., Muniz, A., and Adams, R.B. (2001) Interface: Archaeology and Technology - Digital Archaeology 2001: GIS-Based Excavation Recording in Jordan. The SAA Archaeological Record 1(3):23 - 29. Levy, T.E., Tuttle, C.A., Vincent, M., Howland, M., Richter, A., Petrovic, V., and Vanoni, D. (2013) The 2012 Petra Cyber-Archaeology Cultural Conservation Expedition: Temple of the Winged Lions and Environs, Jordan. Antiquity (Project Gallery) http://antiquity.ac.uk/projgall/levy335/. Levy, T.E, and N.G Smith (2007) On-Site Digital Archaeology: GIS-Based Excavation Recording in Southern Jordan. In Crossing Jordan – North American Contributions to the Archaeology of Jordan. T.E. Levy, M. Daviau, R. Younker, and M. M. Shaer, eds. Pp. 47-58. London: Equinox. Levy, T.E. 2013 Cyber-Archaeology and World Cultural Heritage: Insights from the Holy Land. Bulletin of the American Academy of Arts & Sciences LXVI:26-33. Lowe, D.G. (2004) Distinctive image features from scale invariant keypoints. IJCV, 60(2), pp. 91-110. Snavely, N., S. M. Seitz, R. Szeliski. (2006) Photo Tourism: Exploring image collections in 3D. ACM Transactions on Graphics (Proceedings of SIGGRAPH 2006). Robinson, B. A. (2011) Histories of Peirene: A Corinthian Fountain in Three Millennia. Princeton: American School of Classical Studies at Athens. Rothermel, M., K., Wenzel, D. Fritsch, N. Haala. (2012). SURE: Photogrammetric Surface Reconstruction from Imagery. Proceedings LC3D Workshop, Berlin ,December 2012. Smith, N.G., and T.E. Levy 2012 Real-time 3D archaeological field recording: ArchField, an open-source GIS system pioneered in Jordan. Antiquity 85(331):on-line http://antiquity.ac.uk/projgall/smith331/. Wu, ChangChang. (2007) SiftGPU: A GPU Implementation of Scale Invariant Feature Transform SIFT. http://cs.unc.edu/~ccwu/siftgpu. Wu, Changchang, S. Agarwal, B. Curless, and S. M. Seitz. (2011) Multicore Bundle Adjustment. CVPR 2011, (poster ,supplemental material). © University of the Aegean, 2014, Mediterranean Archaelogy & Archaeometry, 14, 4 (2014) 125-133 Mediterranean Archaeology and Archaeometry, Vol. 14, No 4, pp. 135-141 Copyright © 2014 MAA Printed in Greece. All rights reserved. THE MEDITERRANEAN ARCHAEOLOGICAL NETWORK – A CYBERINFRASTRUCTURE FOR ARCHAEOLOGICAL HERITAGE MANAGEMENT Stephen H. Savage1 and Thomas Levy2 1Archaeological Research Institute, School of Human Evolution & Social Change, Arizona State University, Tempe, Arizona, USA 2Department of Anthropology and Qualcomm Institute, University of California, San Diego, USA Received: 6/12/2013 Accepted: 01/07/2014 Corresponding author: Stephen H. Savage(shsavage@asu.edu) ABSTRACT The Mediterranean Archaeological Network (MedArchNet http://medarchnet.org ) is a series of linked archaeological information nodes. Each node contains a regional database of archaeological sites, sharing a common database structure in order to facilitate rapid information retrieval and display within and across nodes in the network. On the world scene, archaeology produces major new sources of cultural heritage data and material remains that require innovative methods for study, interpretation and public presentation. To take advantage of the growing body of such data, the MedArchNet cyberinfrastructure provides a workable model for researchers from a wide range of fields dealing with cultural heritage to collaborate, discover and monitor resources. In an era of rapidly expanding population and urban development, a system like MedArchNet can provide mechanisms to monitor archaeological site conditions over time and lessen the impact on cultural heritage resources by careful planning, and can significantly enhance site preservation and development potential in the Mediterranean basin. Furthermore, by uniting archaeological site metadata from many disparate datasets and organizations, the MedArchNet cyberinfrastructure dramatically improves the ability of researchers to ask large-scale, cross-regional questions of the archaeological data, providing fresh new insights into one of the most culturally meaningful areas on Earth. KEYWORDS: GIS, Google Earth, Google Maps, Mediterranean, Cyberinfrastructure, archaeological data delivery. 136 STEPHEN H. SAVAGE et al 1. INTRODUCTION The Mediterranean Archaeological Network (MedArchNet) is envisioned as a series of linked archaeological information nodes, each of which contains a regional database of archaeological sites that share a common database structure in order to facilitate rapid query and information retrieval and display within and across nodes in the network [1]. MedArchNet is a signature project of the Center for the Interdisciplinary Study of Art, Architecture and Archaeology (CISA3) at the Qualcomm Institute/California Institute for Telecommunications and Technology (Calit2) at the University of California, San Diego, and the Geo-Archaeological Information Applications (GAIA) Lab, Archaeological Research Institute at Arizona State University. The ultimate vision of MedArchNet (Mediterranean Archaeology Network) is to develop a network of archaeological sites (from remote prehistory to the early 20th century). The MedArchNet website is found at http://medarchnet.org MedArchNet currently contains one active archaeological information node: 1) the Digital Archaeological Atlas of the Holy Land (DAAHL) at http://daahl.ucsd.edu and one node in development -- the Aegean Digital Archaeological Atlas (ADAA), which has yet to be populated and brought online. The data nodes are developed and maintained by the GAIA Lab, and served from Calit2 at UCSD. A prototype web site, shown above, is currently populated by a small database of archaeological sites drawn from the National Geospatial- Intelligence Agency's GEOnet Names Server, located at (http://earthinfo.nga.mil/gns/html/index.html). This prototype is powered by a Google Maps interface, and illustrates the potential for inter-regional archaeological inquiry that can be realized through the MedArchNet concept. 2. A METADATA APPROACH Participants at the recent Workshop on CyberArchaeology, held at the American School of Classical Studies, Athens, on 4 October 2013, discussed the fact that at its most basic level, there is a small set of common variables in all archaeological site databases. These include the site name, the site location (latitude/longitude), time periods when the site is occupied, features that can be found at the site, and links to additional site information, either webbased, or print-based. Participants further expressed the desirability of creating a regional, web-based system of archaeological metadata, encompassing these fields, which would be able to link other, disparate datasets. Essentially, the participants at the workshop were describing MedArchNet in a nutshell. The MedArchNet approach to archaeological site data envisions our data nodes as “switchboards” that contain top-level site and project metadata, plus bibliographic references and extensive use of linked resources outside the MedArchNet data structure. It is not our goal to corral every bit of data about every site in the Mediterrane- © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 135-141 THE MEDITERRANEAN ARCHAEOLOGICAL NETWORK an—an enterprise that would clearly be impossible even if it were desirable. Rather, the MedArchNet approach is designed to let researchers and the public easily find archaeological sites based on location and other attributes such as site type, features, time periods, etc., provide a mechanism for creating substantive maps linked to the various MedArchNet nodes, and then point the user to the locations of substantive research on the site, whether it be on- or offline. The MedArchNet project thus serves to highlight the research of the archaeological community, rather than subsume it under the MedArchNet umbrella; it furthermore fulfills the desire of the Athens Cyberinfrastructure Workshop for a inter-regional metadata system that can unite different regional databases under a single query umbrella, without subsuming any of them. Each data node in MedArchNet maintains a table of data donors, including contact information and primary web sites, and each site contributed by a donor will be “branded” with the donor’s information. Whenever a contributed site is displayed, the donor information is also shown, so the links to the donor’s website are clearly shown, along with specific external resources for individual sites. 3. MEDARCHNET SPONSORSHIP Because the MedArchNet approach to archaeological data maintains and highlights the contributions of individual researchers, university departments, and governmental organizations, it has received the active cooperation and sponsorship of such organizations and individual data contributors. The project actively cooperates with research organizations and government agencies to develop new data nodes and applications. Each of our current nodes has received significant sponsorship. The Digital Archaeological Atlas of the Holy Land (DAAHL) is a sponsored project of the American Schools of Oriental Research, the flagship organization that coordinates North American archaeological research in the Levant. Likewise, the Aegean Digital 137 Archaeological Atlas (ADAA) began as a joint effort of the National Archive of Monuments of the Hellenic Ministry of Culture and the Institute for Aegean Prehistory. As MedArchNet develops additional data nodes, we look forward to expanding our cooperative efforts with additional data donors, research organizations and government agencies. MedArchNet has received sponsorship from the World Universities Network, Equinox Publishing, Inc., Google, Inc., NASA, and the Institute for Aegean Prehistory. 4. “PORTAL SCIENCE” MedArchNet represents the fast growing field of ‘portal science’ and will serve not only researchers and explorers as a platform for international collaboration, but also the general public to share in the excitement of archaeology. The cyberinfrastructure needed to support data collection and representation, data discovery, information integration, and information display for the rich media collections represented by such a network of archaeological data will be developed such that it is extensible to other locations and other archaeological efforts. Thus, for example, our existing data portals could be applied to archaeological sites in Egypt or Turkey, or any other region in the circumMediterranean. In addition, the basic approach that MedArchNet has taken can be easily adapted to other large regions. For example, three prototype websites have recently been developed at the GAIA Lab to showcase the flexibility of the MedArchNet approach, including “The Arabian Archaeological Network” (shown below), “The Sub-Saharan Archaeological Network,” and “The South Asian Archaeological Network.” MedArchNet in general and its data nodes in particular, will serve as the most up-todate source for ‘mining’ stories and narratives of archaeological research in the Mediterranean lands. Archaeological data will © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 135-141 138 STEPHEN H. SAVAGE et al be accessed and displayed over the internet through existing tools such as Google Maps/Google Earth. All of the MedArchNet databases are UTF-8 encoded, so they support multinational character sets; moreover, the Google Translation tool is included at the bottom of every MedArchNet web page, so the output can be translated into any available language at the touch of a button. 5. DATA NODES IN MEDARCHNET As noted earlier, currently MedArchNet consists of one active node, The Digital Archaeological Atlas of the Hoy Land (DAAHL), and one node that is under development, The Aegean Digital Archaeological Atlas. Below, we briefly describe these first two nodes. 5.1 The DAAHL Data Node The first MedArchNet data node created was the Digital Archaeology Atlas of the Holy Land. The DAAHL node brings together many of the developments in information technology that are revolutionizing the fields of archaeology, history, and the social sciences, based on new archaeological discoveries and the latest content concerning one of the most politically complex but meaningful geographic regions in world heritage. The DAAHL node was developed at the GAIA Lab and deployed at Calit2, UCSD; it can be reached at http://daahl.ucsd.edu. The DAAHL project has recently been judged to be one of a small number of “Exemplary Comparative or Thematic Collections” by the Digital Collections Group of the American Schools of Oriental Research Committee for Archaeological Policy (Digital Collections Group 2011). It has been directly linked into the recently developed CAP Projects website at http://asosrtest.org. The DAAHL data node of the MedArchNet system is the only regional, multi-national database of archaeological site and project metadata available for the Levant. It currently contains site metadata for more than 27,000 sites, 60,000 site components, as well as an extensive bibliography, and information related to site conditions. The DAAHL database is a highlycustomized extension of the basic MedArchNet metadata. The web site is organized around two central themes, a series of case studies, the Palestine Exploration Fund Maps, the Virtual Museum and the DAAHL database search and mapping functions. Among its rich data display and query tools are the ability to display the latest archaeological survey and excavation information against an interactive Google Maps background that contains the Palestine Exploration Fund maps—classic archaeological and topographic maps created in the 1870s and 1880s. The PEF maps were scanned and converted into an image pyramid of more than 250,000 tiles for Google Maps display. Site points from the DAAHL database are displayed on top of the historic maps, and each site is back-linked to the database, so that its records can be called up simply by clicking on the site point. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 135-141 THE MEDITERRANEAN ARCHAEOLOGICAL NETWORK Another highly innovative feature is the DAAHL’s Virtual Museum, which displays interactive, 3d objects at their original find locations through a Google Earth API—the user can manipulate the object in all three dimensions as well as the map itself. The most recent development to the DAAHL data node is a mapping function that allows the user to query the database by time period and/or site/feature type. Results are displayed as a series of interac- tive site clusters on the terrain layer of the Google Maps interface. The site clusters are clickable, where doing so zooms to the next level and displays smaller clusters, and individual sites. Site points are also clickable. These maps are suitable for presentation or publication. 5.2 The ADAA Data Node The second node in the MedArchNet cyberinfrastructure is the Aegean Digital Archaeological Atlas, begun through the cooperative effort of the National Archive of 139 Monuments, Hellenic Ministry of Culture and Tourism, the Institute for Aegean Prehistory, and the MedArchNet project. The ADAA project is in its initial stages, with initial data gathering conducted at the Temple University Archaeology Laboratory and at INSTAP Study Center for Eastern Crete at Pachia Ammos, Crete. The Aegean Digital Archaeological Atlas illustrates the “basic” MedArchNet data node. The basic node supports a dedicated database in the MedArchNet format, plus the same kinds of tabular and spatial searches described earlier for the DAAHL website. Like the DAAHL, the ADAA portal supports site Contributor’s listings and “branding” of archaeological sites. The ADAA node also includes online data entry and editing tools. Users with password access to the database have data entry screens available, where site and project records can be created, site pictures uploaded, and other important data entry and maintenance tasks. The site record contains the basic MedArchNet metadata for the site, and is linked to other tables in the database that capture information related to site periods and site/feature types. The combination of period and site/feature type is termed a “Site Component,” and is the preferred method of assigning temporal and archaeological attributes to the site record. Our experience in developing the DAAHL database, MedArchNet, and predecessor archaeological database applications has shown that the most typical query that users make involves a time period and a © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 135-141 140 STEPHEN H. SAVAGE et al site/feature type. For example, a typical query might be “Show me all the Early Bronze Age Dolmen sites,” or ‘Let me see all the Iron II period fortresses.” We encode sites with these combinations so that these kinds of queries can be performed. Simply recording a list of occupation periods and another list of site features does not allow for the kind of query that most users want to conduct. on a site that are dangerous for people or animals, such as the presence of unexploded ordnance, oil or chemical spills, or other factors of this type). Furthermore, the query capabilities that come with each node in the MedArchNet system include a spatial-based query—the user can draw an area on a Google Map window, which might represent an area of potential effect for a planned project. The MedArchNet node can take that drawn area and query its data node for sites that fall in or near it, and the results are returned to the user as a KML file, which can be opened in Google Earth. Each site in the file is back-linked into the database so it can be examined in detail. This allows planners to use the system to interactively design development projects in such a way that they minimize the impact on known sites. 6. SITE CONDITION ASSESSMENT One of the most significant attributes of the MedArchNet approach is the inclusion of a Site Condition Assessment module in each of our data nodes. We recognize that archaeological sites are under tremendous pressure from development, agricultural expansion, warfare, and other factors. These are facets of modern life that have created extremely adverse effects on a finite cultural resource base (see Savage and Rempel, 2013). Too often, though, the tools that are required to help manage and monitor ongoing impacts or threatening developments on sites are lacking. MedArchNet addresses this pressing need by including an assessment module that lets authorized users create what we term a “Visit” record, for any site in the database, and attach to it records related to Site Disturbances 9things already happening on a site that are damaging it), Site Threats (things that have not yet come to pass, but are easily foreseen, such as a road development that will cut through a site in five years’ time), and Site Hazards (conditions 7. FUTURE PLANS FOR MEDARCHNET With the establishment of DAAHL, the basic digital atlas infrastructure for the entire Mediterreanean region is now in place. MedArchNet/DAAHL has been recognized by the American Schools of Oriental Research (ASOR) as an essential tool for addressing «big picture» issues of archaeological dissemination (Levy 2013) and site conservation. Currently, MedArchNet provides a platform for 68 ASOR Committee on Archaeological Policy (CAP) recognized projects to present ‘snap shot’ overviews of their research in Cyprus, Egypt, Israel, Jordan, the Palestinian territories, Turkey. Our goal reach out to these ASOR projects to contribute settlement pattern data from © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 135-141 THE MEDITERRANEAN ARCHAEOLOGICAL NETWORK their research to strengthen DAAHL and initiate new Atlas nodes for the MedArchNet. We would also like to collaborate with other research organizations such as the American School of Classical Studies in Athens, and other non-government organizations that can contribute data to MedArchNet. 8. SUMMARY MedArchNet will have a significant research and education impact by providing easy online access to archaeological data and information. We will deploy the MedArchNet hub and its data nodes, which will provide access to information contributed by each member of the MedArchNet “Virtual Organization”. Via the hub, users will be able to navigate back to the original member sites and databases to access the full information and related data from the 141 respective site. Each data donor receives full recognition and credit for their contribution. Individual portals have been developed initially for ASOR (DAAHL), INSTAP (ADAA), and other partnering groups. The network of linked portals will support collaborations among users and provide a platform for initiating and sustaining discussions related to cross-site thematic areas of study. On the world scene, archaeology produces the major new sources of cultural heritage data and material remains that require innovative methods for study, interpretation and public presentation. To take advantage of the growing body of such data, the MedArchNet cyberinfrastructure will provide a workable model for researchers from a wide range of fields ACKNOWLEDGEMENTS The MedArchNet project has been generously sponsored by the American Schools of Oriental Research, The Institute for Aegean Prehistory, The Worldwide Universities Network, and the support of numerous data contributors. We express our thanks and appreciation to each and every one of them for their support and encouragement. Thanks to INSTAP for a 2013 travel grant that enabled us to present this paper at the Virtual Archaeology conference in Delphi, Greece. FOOTNOTE [1] This paper provides a general description of the project; a more technical description is currently in preparation for submission to Mediterranean Archaeology & Archaeometry near the end of the year. REFERENCES LaBianca O. (2009) A "Big Picture" Research Agenda for ASOR. ASOR Newsletter 59(2):15. Levy TE. (2013) Cyber-Archaeology and World Cultural Heritage: Insights from the Holy Land. Bulletin of the American Academy of Arts & Sciences LXVI:26-33. Savage S.H, and Rempel S.G. (2013) Climate Change and Human Impact on Ancient and Modern Settlements: Identification and Condition Assessment of Archaeological Sites in the Northern Levant from Landsat, ASTER and CORONA Imagery. Final Report, NASA Space Archaeology Solicitation: NNH07ZDA001N-SAP. On-line at: http://gaialab.asu.edu/home/NasaReport.php. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 135-141 Mediterranean Archaeology and Archaeometry, Vol. 14, No 4, pp. 143-153 Copyright © 2014 MAA Printed in Greece. All rights reserved. RECENT ADVANCES IN ARCHAEOLOGICAL PREDICTIVE MODELING FOR ARCHEOLOGICAL RESEARCH AND CULTURAL HERITAGE MANAGEMENT Aikaterini Balla1, Gerasimos Pavlogeorgatos2, Despoina Tsiafakis3 George Pavlidis3 1 31st Ephorate of Prehistoric and Classical Antiquities, Archaeological Museum of Abdera 67061, Abdera, Xanthi, Greece 2 Department of Cultural Technology and Communication, University of the Aegean, Mytilini, Greece 3 Institute for Language and Speech Processing, ATHENA Research Centre, Greece Received: 01/12/2013 Accepted: 17/07/2014 Corresponding author: G.Pavlidis (gpavlid@ceti.gr) ABSTRACT The identification of areas that are insignificant for archaeological research can be used for guidance and support in projects that involve decision-making about the use of land and modern development activities. On the other hand, the identification of areas significant for archaeological research can contribute to archaeological knowledge and minimise the risk of unsuccessful excavations. This paper presents a review of the most recent and representative applications of predictive modelling in Archaeology, which demonstrate that predictive models can be successfully exploited by archaeological research and Cultural Heritage Management (CHM). KEYWORDS: archaeological predictive modelling, Cultural Heritage management, archaeological research 144 AIKATERINI BALLA et al 1. INTRODUCTION The primal objective of Archaeology is the composition of the history and the understanding of past cultures through the study and interpretation of the natural relations of archaeological finds and the ideological context within which they operate. The problem, however, is always more complex in practice. On one hand, the discovery of archaeological remains is coincidental and mainly a result of modern development interventions, which, however, lead to partial or total destruction of archaeological sites. On the other hand, a number of reasons, such as the lack of financial resources ensue lack of systematic archaeological excavations and therefore, incomplete knowledge of the archaeological remains in many areas. Even in the cases of more systematic excavations, the studied areas are usually necessarily small and, consequently, the archaeological information collected and studied cannot be easily compared or opposed to data from other areas related to the same human activity, from which these archaeological remains were generated. The effort to address these problems led to the development and implementation of methodologies that would be able to recognize and identify possible areas of human activity and use in the past. Within this research context, the use of archaeological predictive modelling in the past years has yielded important expertise that can be used successfully both in CHM and archaeological research. Management of cultural heritage over the recent decades received more attention and resources in order to face threats related to the physical damage of cultural assets during modern development activities (urban development, large-scale agriculture and mining), or even threats like looting and environmental threats like erosion (Neumann and Sanford 2001). Recently, the identification and protection of cultural sites has been a subject of legislation that in many cases criminalised land development prior to conducting a cultural resources survey to identify any cultural sites that may be affected (Neumann and Sanford 2001). Nowadays, CHM can be significantly empowered by the application of scientific approaches and digital technologies that provide fast and reasonably accurate prediction of the existence of cultural sites prior to development projects. Predictive Modelling (PM), in particular, has already been proven as a valuable tool for the rescue of archaeological remains and archaeological data that would otherwise have been lost due to modern development. Additionally, in terms of archaeological research, PM has already been successfully exploited by archaeologists, and is further expected to be an integral part of archaeological practice, in interpreting and understanding the socio-economic structure of the past. The identification of new archaeological sites through the application of PM techniques would enrich archaeological knowledge about ancient culture and would contribute to the study of ancient topography, as the discovery of new sites can result in finding yet undiscovered areas of archaeological interest. Furthermore, the predictive models can be used as an efficient solution to the lack of funding, by providing insight on the existence of archaeological remains in studied areas, thus minimising the need for trial excavations. 2. ARCHAEOLOGICAL PREDICTIVE MODELLING The most commonly used definition of PM in Archaeology belongs to Kohler & Parker (1986), who describe it “a technique that, at a minimum, tries to predict the locations of archaeological sites or materials in a region, based either on a sample of that region or on fundamental notions concerning human behaviour”. According to Verhagen (2007), PM is based on the assumption that the location of archaeological sites is not random, but it is associated to specific characteristics of the natural environment and factors related to human activity and human behavioural norms in the past. By identifying this causal relationship between certain environ- © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 143-153 RECENT ADVANCES IN ARCHAEOLOGICAL PREDICTIVE MODELING mental and geographical characteristics and known archaeological site locations, repeating patterns can be identified, creating a statistical model that can be applied to unsurveyed areas in order to identify new locations that may also have been occupied by similar human activities. Namely, PM can be conceptualised as a specialised form of location-allocation analysis, where the aim is to allocate suitable locations to specific types of human activity and their archaeological remains (Van Leusen 2002). The data used to create an archaeological predictive model always arise from the relationship of archaeological sites with the natural and cultural environment. It is clear, however, that the input parameters of an archaeological predictive model should be associated both with the study area and the subject of study. Jaroslaw and HildebrandtRadke (2009) for example, report that many studies, which examine the locational processes of ancient settlements (both before and after the introduction of GIS techniques to Archaeology), suggest that, apart from socioeconomic factors, features such as topographic relief, distance from water bodies or soil cover type, had also an important role (Bauer et al. 2004, Duke 2003, Fletcher 2008, Kvamme 1992, Stancic and Kvamme 1999, Warren 1990, Willey 1953, Williams 1956, Williams et al. 1973). Those features, however, cannot be used as input data on predictive models for other types of archaeological sites (for example burial mounds or sanctuaries). Therefore, in any case, it is necessary to study thoroughly the particular type of archaeological site and extract the criteria that led to the specific human decision rules. It is clear that those factors-criteria can vary even for the same type of archaeological site, as they may be related to a specific time period, region or specific cultures. 3. CONTRIBUTION TO CHM AND ARCHAEOLOGICAL RESEARCH: INDICATIVE CASE STUDIES The increased attention in archaeological PM led to an extensive literature research and numerous case studies. In this study, 145 we present indicative studies of archaeological predictive modelling of the past ten years, analyse their aims and scopes to both CHM and archaeological-academic research and present their experimental results. Siart et al. (2008) conducted geospatial analyses of archaeological sites and communication paths of the Bronze Age (Minoan Neo-Palace Period - about 1650 BC) in the region of Mount Ida in central Crete, for the detection of Bronze Age infrastructures and potential archaeological candidate sites. The study included the development of an information system which visualized the main geological characteristics of Mount Ida, the mapping (based on field research) of the geomorphology, the vegetation, the hydrology and known archaeological sites of the study area, remote sensing techniques, least cost analyses, predictive modelling and GIS. The researchers stress the need to include in archaeological analyses comprehensive sets of environmental variables that might have influenced ancient settlement patterns and show the advantages of using a multimethod approach for reconstructing ancient landscapes. The study can be used, according to the researchers, for unexplored areas where the archaeological data are poor and the environmental conditions adverse and can contribute to the acquisition of new archaeological knowledge. Vaughn and Crawford (2009) used GIS in conjunction with remotely sensed imagery, paper map data, and Binary Logistic Regression to predict the probability of ancient Maya archaeological site presence in Northwest Belize in Central America. The input variables in the modelling process were selected using Binary Logistic Regression and were associated with both the representation of the ancient landscape and the current environmental terrain. The optimal function of the model was achieved by the use of the following variables: vegetation index (Tasselled Cap Greenness index), viewshed (eastern aspect), proximity to flat land. The model suggested that there is a higher probability of settlements’ oc- © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 143-153 146 AIKATERINI BALLA et al currence in locations oriented to the east, with easy access to arable land and high vegetation index. The evaluation of the model was examined using the Kvamme predictive gain G, which generated a moderate gain statistic of 0.26. The proposed methodology can be used, according to the researchers, to the archaeological-academic research regarding the Mayan culture, and moreover, greatly reduce the cost and time required for future field research. Fernandes et al. (2011) used statistical methods and GIS to study the settlement site in Malia in the ProtoPalatial period and understand the causal relationship between settlement locations and independent variables. They applied two predictive models: a purely environmental, which used as criteria-variables altitude, distance from sea coast, geomorphology, soil depth and density of the water bodies and a mixed model (environmental-historical), which used, apart from most of the environmental variables mentioned above, criteria concerning human factors such as the major urban centres of the time. The researchers introduced in the modelling process an algorithm in order to identify the best performance of the model. The testing and evaluation of the results of the two models ascribed the best predictive ability in the environmental-historical model. Graves (2011) developed and applied two predictive models in order to identify human settlement and occupation activity in the mainland of Scotland during the Neolithic Period. The study was based on GIS, statistical methods and an inbuilt presumption that locations of settlement or occupation activity on the mainland were related to the locations of the timber halls, pits, and chambered cairns. A GIS was used to extract environmental variables commonly used in archaeological predictive modelling from input sites and non-site locations: the variables included elevation, slope, aspect, local relief, distance to the nearest source of water, cost-distance to the nearest source of water, and viewshed. An important conclusion of the study was that the variables of viewshed and proximity to water bodies seem to have greater significance in the selection of the sites in relation to the other criteria as they were identified as powerful predictors. To evaluate the prediction results Graves used gain G which showed that, out of a total of 74 ‘activity’ sites, models 1 and 2 can successfully predict 86% and 84.5%, respectively. However, as Graves points out, without fieldwork it is impossible to know the real gain of each model and therefore the gains should be treated as preliminary and secondary to field tests. The researcher finally fosters the hope that in the future, the proposed models could test in the field the archaeological theories about the perceived relationship of the input sites to settlement or occupation activities. The study subject of Aubry et al. (2012) was the open-air rock art of the late Ice Age and the Iron Age, which shows a similar spatial distribution along the rivers Côa and Douro (Portugal) and orientation towards Southeast. The researchers tried to determine whether the artists of the two periods deliberately chose the same natural environments for their art, or if the current spatial distribution of their art remains resulted from other processes formations or corrosion of rocks, before or after the artistic formations. The study included analysis of geological structures of local and regional level, field measurements, analysis of the hydrological network, etc., whereby the researchers concluded that the distribution of the remains of prehistoric art is the result of natural processes and geological formations combined with different conservation/erosion of the rocks’ surface. The factors that affected, according to the researchers, the erosion of these surfaces were the diversity of solar radiation, humidity, and the growth of algae and lichen. The researchers combined the interpretation of observations from field research with probabilistic processes, pairwise comparisons of observed patterns of the two time periods and geospatial analysis through GIS to develop predictive models, which would identify areas with similar geological and climatic characteristics. The © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 143-153 RECENT ADVANCES IN ARCHAEOLOGICAL PREDICTIVE MODELING archaeological input data (rock art occurrences) were used to evaluate the predictive models and external validation maps, with the results showing an agreement of 70%-80%. Most importantly, the following field survey, revealed unknown rock paintings in areas with high and very high values. The researchers conclude that the predictive model would serve as a useful tool in archaeological fieldwork and Cultural Resource Management. Luczak (2013) used two regression modelling methods: Generalized Linear Model (GLM) and Generalized Additive Model (GAM) and examined their ability to analyse and predict archaeological sites locations in southern Poland. The archaeological datasets used in the modelling process came from the field survey record stored in the database of the Polish Archaeological Record (PAR) and represented sites from two different periods: Neolithic and Medieval (Early and Late). Through typical GIS software and procedures 11 environmental variables (representing hydro-morphological terrain attributes and soil types) and a set of cultural variables (visibility and distance from political and administrative centres (castle, fortress or fortified settlement) were obtained and statistical analysis (e.g. density plots, correlation coefficients, etc.) was used in order to determine past settlement preferences, their potential influence on site location and also to examine the differences between settlement patterns in these periods. The models’ predictive ability was evaluated through the ROC curve function (AUC), which showed that statistically the GAM models give better predictions than the GLM models. Łuczak chose the best modelling method (the GAM models) to produce binary probability (0-1) maps, which were next used to create 2 final maps for Neolithic and Medieval periods combining permanently and temporarily settled sites predictions. The researcher concluded that predictive models could serve as a great tool for archaeologists in settlement research, but also stresses the need to be verified and checked for their reliability, apart from their statistical eval- 147 uation, through field surveys. Moreover, it is necessary to use a variety of prediction methods to understand and interpret different results and also examine the accuracy of the models not only for chronologically different sites (prehistoric and historical), but also for different site types (e.g. temporarily settled, permanently settled, hunter shelter, monuments etc.). Verhagen et al. (2013), noting the relative absence of socio-cultural factors in prehistoric and historical site location choice attempted to address the unexplored human factor in predictive modelling, by developing a protocol using both environmental and socio-cultural factors that can easily be implemented for different regions and time periods. The development of the predictive model was based on cross-regional comparisons of settlement location factors, like slope, aspect and solar radiation made in the 1990s by analysing the environmental context of Roman settlements in the French Rhône Valley. However, for the current study, the researchers expanded the set of variables with “socio-cultural” factors, in particular accessibility, visibility, and the effect of previous occupation in order to establish whether including socio-cultural factors actually made a difference for the interpretation of site location patterns and predictive model quality. Though the prediction of settlement locations was implied, the researchers stressed that the optimal model performance was not the main goal of their study, as would be the goal of standard statistical approaches like logistic regression. Instead, the “non-performance” of a variable was considered an equally important result, as the protocol’s aim was to extract the main factors that influenced settlement location over a longer term. In conclusion, the study, though preliminary, showed that there were limitations to both environmental and certain social variables, as they may not be relevant for other archaeological settings, or cannot even be modelled in all situations because of poor available archaeological data. An extensive review of the related literature indicates the lack of applications relat- © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 143-153 148 AIKATERINI BALLA et al ed to cemeteries and burial sites. Undoubtedly, burial mounds, tombs and cemeteries have been the subject in many studies, which, however, examine the correlation between topography and their location on the landscape (De Reu et al. 2011, Löwenborg 2010b), chronological estimations (Löwenborg 2009), viewshed and visibility (Fisher et al. 1997, Lageras 2002, Wheatley 1995, Woodman 2000) or simply included among other archaeological data, the locations of funerary monuments and cemeteries to map archaeological sites. The studies found in literature regarding exclusively the prediction of burial monuments or mounds are rare and will be presented shortly in the following paragraphs. Al-Muheisen & Al-Shorman (2004) used GIS to analyse the landscape and the mortuary practices in three cemeteries of the late Roman and early Byzantine period in the region Bediyeh (North Jordan), in order to obtain the spatial relationships of the various features at the site in a ritual and cultural context and, thus, reconstruct past behavioural practice and cult. Within the framework of their research they developed an inductive predictive model for burial monuments and applied it to the archaeological data of the western cemetery, which, based on the typology of the tombs (chambered) and the funerary gifts found inside the tombs, was believed to be predestined for the dead of the higher social classes. In the model building process they used three variables (proximity between the funerary monuments, slope, viewshed), to which a different weight was assigned, based on the frequency of the known monuments in relation to each of these criteria. The prediction results were evaluated with the gain G, attaining 0.82 value for the best performance of the model. It was noted that the locations indicated by the model as the most probable for tomb occurrence were places prominent and visible to other tombs, which can be attributed, as the researchers speculate, to the higher social class of the deceased and their relatives’ desire for their tombs to be visible. Apart from the contribution in the understanding of the various spatial relationships among the various features of the site and interpretation of the archaeological data, the predictive model can also be used as a guide and, moreover, as a “cost reducing” tool for future excavations. Fry et al. (2004) within the context of Cultural Heritage Management and protection in Norway suggested a methodology for the development of predictive models that would be able to identify possible locations of burial monuments’ of the Bronze and the Iron Age. For the spatial analysis and the mapping of the spatial data they used GIS, readily available environmental data and visual analysis. The predictive model indicated areas of high probability of burial mounds’ existence and successfully provided 94% of the known mounds in areas that cover only 12% of the total survey area. The results led to the discovery of new sites of archaeological interest and have contributed significantly to the understanding of the tomb distribution in Norway. Moreover, the generated maps would serve as a useful tool for heritage managers and development projects’ planners by identifying areas where development may risk damaging antiquities. Burns et al. (2008) have created models that predict possible locations of funerary monuments in the Theban necropolis in Luxor, Egypt, and examined its usefulness in understanding the reasons that led to the preference of those locations by the ancient Egyptians. The location of the tombs was examined in relation to geology, slope, elevation, fractures, and religious-funerary practices (orientation of tombs, proximity to temples). The environmental and the archaeological data were quantified using GIS and statistical analyses and a predictive model was developed, which could be linked to the database of the Egyptian Antiquities Information System (EAIS) created by the Egyptian government for the protection of Cultural heritage in order to indicate which sites should be avoided within the context of modern development planning or further studied in terms of archaeological research. The model results showed © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 143-153 RECENT ADVANCES IN ARCHAEOLOGICAL PREDICTIVE MODELING that the stable soil, the eastward orientation and the orientation to the royal temples were important factors in the decisionmaking process regarding the siting of the tombs. These models, however, did not allow determining the degree of importance of these factors. On the other hand, it seems that the criteria of elevation, slope and fractures did not affect significantly the choice of the tombs’ location. However, according to the researchers, this does not mean that the ancient Egyptians did not take them under consideration, but that their study obviously was lacking relevant archaeological documentation. The researchers suggested for future study the examination of the role of other factors that may have affected the tombs’ location. The most important outcome of their study is related to the lack of archaeological knowledge about the reasons-factors that led to the locational selection of the tomb construction. Most archaeological studies are limited to the discovery of the identity of the deceased and not necessarily to the knowledge of the criteria, which led to the construction of the burial monument at this particular site. Further investigation of this field of Egyptian Archaeology could lead to better models and thus to the discovery of new tombs and the understanding of the complexity of decision making in the past. Balla et al. (2013) created a predictive model for the detection of Macedonian tombs in Northern Greece. The proposed methodology was based on the following procedures: through archaeological research and data aggregation, assumptions related to the location of the sites of interest were formulated, resulting in the selection of criteria considered to have influenced the siting of the Macedonian tombs. Thus, by taking under consideration the literature research on all Macedonian tombs, and, also, based on the existing geographic data, the researchers ended up with four environmental (altitude, slope, soil hardness, distance from rivers) and two cultural parameters (distance from settlements, dis- 149 tance from roads). At the core of the proposed methodology, a multi-criteria analysis on geospatial data processing technologies (GIS), predictive modelling techniques and fuzzy logic was applied to the study area in order to create a predictive model that would be able to provide map regions assigned with specified probability of Macedonian tombs’ occurrence. The model was created and tested under various combinations of parameters related to the criteria. The results were evaluated by using a commonly used predictive gain, which proved the efficiency of the model’s predictive ability in providing answers to a series of questions related to the problem at hand and could benefit both archaeological research (discovery and study of new tombs, contribution in ancient topography etc.) and cultural heritage management and protection. On one hand, in terms of archaeological research, the model provided very promising results, identifying a high percentage of known Macedonian tombs within relatively small spatial zones. Namely the model identified a total of 87.95% of the known Macedonian tombs within a 16.55% of the total survey area and in the case of the “very high probability” areas, identified 55.42% of the known Macedonian tombs in an area smaller than 6% of the total surface area (namely in the 1/19 of the total survey area). On the other hand, the results produced by the proposed method were considered of great importance for cultural heritage protection. In that case, where the aim was the identification and knowledge of large areas that contain no, or the least possible, archaeological sites, the model indicated a large area with total Macedonian tombs’ absence (31.73% of the total survey area), which could be excluded in development planning. Table I summarises the recent works in archaeological predictive modeling presented in this paper. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 143-153 150 AIKATERINI BALLA et al Table I. Summary of recent works in archaeological predictive modeling Authors Siart et al. (2008) Prediction Aim Bronze Age infrastructures Area/period of interest Region of Mount Ida in central Crete (Bronze Age) Northwest Belize in Central America Tools Remote sensing, GIS Vaughn and Crawford (2009) Ancient Maya archaeological site presence Fernandes et al. (2011) Causal relationship between settlement locations and environmental and cultural variables Human settlement and occupation activity (Neolithic Period) Malia, Crete (ProtoPalatial period) Statistical methods and GIS Mainland of Scotland GIS, statistical methods and specific limitations Aubry et al. (2012) Factors that influenced the location of openair rock art in different time periods Rivers Côa and Douro, Portugal (late Ice Age and the Iron Age) Probabilistic processes, pairwise comparisons and geospatial analysis through GIS Luczak (2013) Identification of Neolithic and Medieval (Early and Late) settlements Protocol using environmental and sociocultural factors that can easily be implemented for different regions and time periods Burial sites identification Southern Poland (Neolithic and Medieval Early and Late) Generalized Linear Model (GLM) and Generalized Additive Model (GAM) --- Region Bediyeh (North Jordan) GIS Fry et al. (2004) Burial sites identification Norway (Bronze and the Iron Age) GIS, readily available environmental data and visual analysis Burns et al. (2008) Burial sites identification Theban necropolis in Luxor, Egypt GIS and statistical analysis Balla et al. (2013) Burial sites identification (Macedonian Tombs) North Greece (Late Classical and Hellenistic) Multi-criteria analysis GIS and fuzzy logic Graves (2011) Verhagen et al. (2013) Al-Muheisen & Al-Shorman (2004) Roman settlements in the French Rhône Valley 4. CONCLUSIONS Predictive models for archaeological sites have become an integral part of archaeological applications, displaying an increasing number of methodologies that attempt to meet different purposes and needs of Archaeology (contribution to archaeological research or CHM). Despite the differences in the analysis approach, practically, the same process is followed in all GIS, remote sensing, paper map data, binary logistic regression Outcome Contribution to archaeological research and CHM Kwamme predictive gain 0.26, discovery of new sites, contribution to archaeological research Better performance by using environmentalhistorical model Successfully predicted 86% of 74 sites, contribution to archaeological research 70%-80% prediction accuracy and new sites identification, contribution to archaeological research New sites identification, contribution to archaeological research Extracted the main factors that influenced settlement location Kwamme predictive gain 0,82, discovery of new sites, cost reduction, contribution to archaeological research 94% accuracy in 14% of total survey area, discovery of new sites, contribution to archaeological research Contribution to archaeological research and CHM Probability maps, contribution to archaeological research and CHM cases: their creation is based on the correlation of environmental and cultural parameters with known archaeological sites. The statistical analysis of those archaeological sites correlates the spatial variability of the environmental and cultural parameters with other sites of possible archaeological interest, based on specific decision making rules. Predictive models can be used in archaeological-academic research by indicating © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 143-153 RECENT ADVANCES IN ARCHAEOLOGICAL PREDICTIVE MODELING areas of high probability to find sites of archaeological interest and therefore need further investigation. Thus, the discovery of new archaeological sites would certainly add new data to the existing archaeological knowledge and the study of the historical topography, providing a clearer picture of the number of sites of human activity in the past, their spatial relationships, their connecting networks (roads) etc. Additionally, they can contribute to a cost reduction by minimising the requirements for trial excavations. On the other hand predictive models can be used for cultural heritage protection, 151 where the aim is to identify the areas that do not include sites of archaeological interest and, thus, exclude them from any development/urban planning. When used as a spatial guidance and support for projects of land use and modern development, predictive models can prevent possible future damage of archaeological sites. In conclusion, predictive modelling can be a successful tool in archaeological analyses and studies with the potential to give new impetus to archaeological thinking and interpretation of the remains of the past. REFERENCES Al-Muheisen, Z., Al-Shorman, A. (2004) The Archaeological Site of Bediyeh: the Constructed Landscape. Syria 81, 177-190. Aubry, T., Luís, L., Dimuccio, A. L. (2012) Nature vs. Culture: present-day spatial distribution and preservation of open-air rock art in the Côa and Douro River Valleys (Portugal). Journal of Archaeological Science 39:4, 848-866. Balla, A., Pavlogeorgatos, G, Tsiafakis, D., Pavlidis, G. (2013) Locating Macedonian tombs using predictive modeling. Journal of Cultural Heritage 14:5, 403–410 Bauer, A., Kathleen, N., Park, L., Matney, T. (2004) Archaeological site distribution by geomorphic setting in the southern Lower Cuyahoga River Valley, Northeastern Ohio: initial observations from a GIS database. Geoarchaeology: An International Journal 19:8, 711–729. Burns, G., Fronabarger, A. K., Whitley, T. G. (2008) Predictive modeling of cultural resources in the Theban Necropolis, Luxor, Egypt. In: Posluschny, A., Lambers, K., Herzog, I. (Eds.), Layers of perception. Proceedings of CAA Conference, 35th Annual Meeting, Berlin, Germany 2007, Rudolf Habelt GmbH, Bonn. De Reu, J., Bourgeois, J., De Smedt, P., Zwertvaegher, A., Antrop, M., Bats, M., De Maeyer, P., Finke, P., Van Meirvenne, M., Verniers, J., Crombé, P. (2011) Measuring the relative topographic position of archaeological sites in the landscape, a case study on the Bronze Age barrows in northwest Belgium. Journal of Archaeological Science 38:12, 3435-3446. Duke, C. (2003) Quantifying Palaeolithic landscapes: computer approaches to terrain analysis and visualization. In: Doerr M., Sarris A. (Eds.), The digital heritage of archaeology. Proceedings of CAA Conference, 29th Annual Meeting, Heraklion, Crete, Greece, 2002, Hellenic Ministry of Culture, Athens, Greece, 2003, pp. 139–146. Fernandes, R., Geeven, G., Soetens, S., Klontza-Jaklova, V. (2011) Deletion, Substitution, Addition (DSA) model selection algorithm applied to the study of archaeological settlement patterning. Journal of Archaeological Science 38:9, 2293-2300. Fisher, P., Farrelly, C., Maddocks, A., Ruggles, C. (1997) Spatial Analysis of Visible Areas from the Bronze Age Cairns of Mull. Journal of Archaeological Science 24:7, 581-592. Fletcher, R., (2008) Some spatial analyses of Chalcolithic settlement in Southern Israel. Journal of Archaeological Science 35:7, 2048–2058. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 143-153 152 AIKATERINI BALLA et al Fry, G. L. A., Skar, B., Jerpansen, G., Bakkestuen, V., Erikstad, L. (2004) Locating archaeological sites in the landscape: a hierarchical approach based on landscape indicators. Landscape and Urban Planning 67:1-4, 97-107. Graves, D. (2011) The use of predictive modelling to target Neolithic settlement and occupation activity in mainland Scotland. Journal of Archaeological Science 38:3, 633–656. Jaroslaw, J. and Hildebrandt-Radke, I. (2009) Using multivariate statistics and fuzzy logic system to analyse settlement preferences in lowland areas of the temperate zone: an example from the Polish Lowlands. Journal of Archaeological Science 36:10, 2096. Kohler, T. A. and Parker, S. C. (1986) Predictive models for archaeological resource location. In: Schiffer, M. B. (Ed.), Advances in Archaeological Method and Theory 9. Academic Press, New York, pp. 397-452. Kvamme, K.L. (1992) A predictive site location model on the high plains: an example with an independent test. Plains Anthropologist 37:138, 19–40. Lageras, K. E. (2002) Visible intentions? Viewshed analysis of Bronze Age burial mounds in western Scania, Sweden. In: Scarre, C. (Ed.), Monuments and Landscape in Atlantic Europe. Perception and Society during the Neolithic and Early Bronze Age. Routledge, London/New York, pp. 179-191. Löwenborg, D. (2009) Landscapes of death: GIS modeling of a dated sequence of prehistoric cemeteries in Västmanland, Sweden. Antiquity 83:322, 1134-1143. Löwenborg, D. (2010) Digital Perceptions of the Landscape: A GIS based analysis of the location of burial grounds in Västmanland, Sweden. Acta Archaeologica 81:1, 124-137 Luczak, A. (2013) Using predictive modeling methods as a way of examining past settlement patterns: an example from southern Poland. Colloquium. Non-destructive approaches to complex archaeological sites in Europe: a round up. Ghent University, 1517/1/2013, Belgium. Neumann, T.W. and Sanford, R.M. (2001) Practicing Archaeology: A Training Manual for Cultural Resources Archaeology. Rowman and Littlefield Pub Inc, Indiana University. Siart, C., Eitel, B., Panagiotopoulos, D. (2008) Investigation of past archaeological landscapes using remote sensing and GIS: a multi-method case study from Mount Ida, Crete. Journal of Archaeological Science 35:11, 2918-2926. Stancic, Z., Kvamme, K. (1999) Settlement pattern modelling through Boolean Overlays of social and environmental variables. In: Barceló, J. A., Briz, I., Vila, (Eds.), New Techniques for Old Times, Proceedings of CAA Conference, 26th Annual Meeting, Barcelona, 1998, BAR International Series 757, Archaeopress, Oxford, pp. 231–237. Van Leusen, P.M. (2002) Pattern to process: methodological investigations into the formation and interpretation of spatial patterns in archaeological landscapes, Ph.D. thesis, University Groningen, Groningen. Vaughn, S. and Crawford, T. (2009) A predictive model of archaeological potential: An example from northwestern Belize. Applied Geography 29:4, 542-555. Verhagen, P. (2007) Case Studies in Archaeological Predictive Modelling. Leiden University Press. Verhagen, P., Nuninger, L. Tourneux, F.-P., Bertoncello, F. and Jeneson, K. (2013) Introducing the human factor in predictive modelling: a work in progress. In: Earl, G., Sly, T., Chrysanthi, A., Murrieta-Flores, P., Papadopoulos, C., Romanowska I. and Wheatley, D. (Eds). Archaeology in the Digital Era. Papers from the 40th Annual Conference of Computer Applications and Quantitative Methods in Archaeology (CAA), Southampton, 26-29 March 2012, pp. 379-388. Warren, R.E. (1990) Predictive Modelling in Archaeology: A Primer. In: Allen, K. M. S., Green, S. W., Zubrow, E. B. W. (Eds.)., Interpreting Space: GIS and Archaeology. Taylor & Francis, London, pp. 90–111. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 143-153 RECENT ADVANCES IN ARCHAEOLOGICAL PREDICTIVE MODELING 153 Willey, G.R. (1953) Prehistoric Settlement Patterns in the Viru Valley, Peru. Bureau of American Ethnology Bulletin 155. Smithsonian Institution, Washington, DC. Williams, S. (1956) Settlement patterns in the lower Mississippi valley. In: Willey, G. (Ed.), Prehistoric Settlement Patterns in the New World. Viking Fund Publications in Anthropology 23, New York, pp. 52–62. Williams, L., Thomas, D., Bettinger, R. (1973) Notions to numbers: Great Basin settlements as polythetic sets. In: Redman, C.L. (Ed.), Research and Theory in Current Archaeology. John Wiley & Sons, New York, pp. 215–237. Wheatley, D. (1995) Cumulative Viewshed Analysis: a GIS-based method for investigating intervisibility, and its archaeological application. In: Lock, G., Stancic, Z. (Eds.), Archaeology and Geographical Information Systems: A European Perspective. Taylor & Francis, London, pp. 171-186. Woodman, P. E. (2000) A predictive model for Mesolithic site location on Islay using logistic regression and GIS. In: Mithen, S. J. (Ed.), Hunter-Gatherer Landscape Archaeology: The Southern Hebrides Mesolithic Project, 1988-98, Archaeological Fieldwork on Colonsay, Computer Modelling, Experimental Archaeology, and Final Interpretations. The McDonald Institute for Archaeological Research, Cambridge, pp. 445-464. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 143-153 Mediterranean Archaeology and Archaeometry, Vol. 14, No 4, pp. 155-164 Copyright © 2014 MAA Printed in Greece. All rights reserved. 3D VIRTUAL RECONSTRUCTION OF ARCHAEOLOGICAL MONUMENTS Andreas Georgopoulos Laboratory of Photogrammetry, National Technical University of Athens, Athens, Greece Received: 01/12/2013 Accepted: 29/06/2014 Corresponding author: A. Georgopoulos (drag@central.ntua.gr) ABSTRACT 3D Virtual Models are the future of the representation of the existing and destroyed architectural heritage. The term reconstruction defines the re-building of a monument to its state at the time of its history chosen for that particular representation. In recent years the evolution of the technology, has contributed significantly in many aspects of the field of cultural heritage preservation and recording. Techniques like digital image processing, digital orthophoto production, terrestrial laser scanning and 3D model processing have enabled the production of such alternative products. In this paper two characteristic cases of 3D virtual reconstruction of non-existing monuments are presented: The Middle Stoa in the Athens Agora and the Church of San Prudencio’s Monastery in Spain. All data collected were evaluated and used appropriately for the final products. It is evident that the data collected do not all belong to the target periods and not all the data necessary to built up the models are available today. Therefore, one needs to carefully select the data corresponding to the period of study and complete them with suitable hypotheses. It is imperative that both tasks must be done in collaboration with archaeologists and architects. In this context a data hierarchy was developed, based on their reliability and correctness. The data were categorized for their reliability after careful evaluation their accuracy depending on the source. In this paper a 'Reliability' matrix for creation of digital models for cultural heritage research is presented. Sometimes the data appear in more than one source; in this case they must be checked for correspondence. All different sources should be evaluated and used accordingly for the final product. The procedures followed are briefly described and the results are presented and assessed for their reliability and usefulness. KEYWORDS: Virtual Reconstruction, 3D models, Digital Cultural Heritage, ICT tools 156 ANDREAS GEORGOPOULOS 1. INTRODUCTION Nowadays, the rapid advances of Digital Technologies also referred to as Information Communication Technology (ICT), have provided scientists with new powerful tools. Especially in the field of Cultural Heritage Documentation, we are now able to acquire, store, process, manage and present any kind of information in digital form. Nowadays this may be done faster, more accurately and more completely and in this way a larger base of interested individuals is built that this information may reach. The use of Digital Technologies in preservation and curation in general of cultural heritage is also mandated by UNESCO. With the Charter on the Preservation of the Digital Cultural Heritage (http://portal.unesco.org/en/ev.phpURL_ID=17721&URL_DO=DO_TOPIC&U RL_SECTION=201.html) this global organization proclaims the basic principles of Digital Cultural Heritage for all civilized countries of the world. At the same time numerous international efforts are underway with the scope to digitize all aspects of Cultural heritage, be it large monuments, or tangible artefacts or even intangible articles of the world’s legacy. 2. MOTIVATION The involvement of contemporary Digital Technologies (ICT) in the domain of Cultural Heritage has increased the gap between Providers, i.e. those who master these techniques and are able to apply them and the Users, i.e. those scholars traditionally concerned with the Cultural Heritage. This gap was caused mainly due to the mistrust of the latter towards contemporary technologies and lately ICT. However systematic efforts have been applied, like RecorDIM by CIPA, i.e. the International Committee of Architectural Photogrammetry formed by ICOMOS and ISPRS (International Society for Photogrammetry and Remote Sensing (http://cipa.icomos.org/index. php?id=43) which have managed to narrow if not bridge this gap. This current effort concerned with the 3D virtual reconstruction of monuments is motivated exactly by this endeavour to bridge this gap. This will only be done through deep understanding of each other’s needs and through proper exploitation of ICT with the benefit of Cultural Heritage always in mind. In addition, the notion of virtual reconstruction is introduced and its use for bringing the reconstructed monuments into a museum environment is investigated. 3. ICT TOOLS 3.1 Definition The available contemporary digital tools include mainly instrumentation for digital data acquisition, such as terrestrial laser scanners, structured light scanners, digital optical, thermal and range cameras, digital total stations etc., software for processing and managing the collected data, such as structure from motion (SfM) and -of course- computer hardware, for running the demanding software, storing the data and presenting them in various forms. 3.2 Impact Already the introduction of digital technologies has altered the way we perceive fundamental notions like indigenous, artefact, heritage, 3D space, ecology etc. At the same time they tend to transform the traditional work of archaeologists and museums as they are known so far. In other words Digital Technologies redefine the relationship to Cultural Heritage, as they enable universal access to it and they also connect traditional cultural institutions to new “audiences”. Finally traditional museums and archaeological sites through the use of contemporary technologies appeal to new generations, as the latter are, by default, computer literate. 3.3 New products, possibilities and uses The impact of digital technologies to the domain of Cultural Heritage has enhanced speed and automation of the procedures which involve processing of the digital © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 155-164 3D VIRTUAL RECONSTRUCTION OF ARCHAEOLOGICAL MONUMENTS data and presentation of the results. At the same time accuracy and reliability has been substantially enhanced. However, most important is the ability to provide to the users new and alternative products, which include two dimensional and three dimensional products, such as orthophotos and 3D models. All in all the digitization of the world’s Cultural Heritage, be it tangible or intangible is now possible. 3D modelling, on the other hand, is the process of virtually constructing the threedimensional representation of an object. The use of 3D models is highly increased nowadays in many aspects of everyday life (cinema, advertisements, museums, medicine etc). This paper focuses on their use for representing, reviving, interacting and studying Cultural Heritage in an interactive way. 3.4 3D Digital Data Digital data acquisition is nowadays performed with (a) geodetic digital total stations, which produce 3D coordinates of single points in space, (b) digital image processing, which produces 2D or 3D products and (c) with digital scanning devices, which produce 3D point clouds of the objects. The common attribute of the above is the three dimensional data acquisition, which enables the development of 3D virtual models. It is these 3D models that have made the 3D virtual reconstructions possible. 3D models can be simple linear vector models or they can consist of complex textured surfaces depending on the object and their final use. As the specific technology advanced, 3D models were used for multiple purposes. Initially they simply served as means for visualization. Gradually, however, they contributed to other uses, such as study, description purposes and restoration interventions and lately for virtual reconstruction and engineering applications (Valanis et al. 2009). 157 Technological advances have provided 3D modelling software with numerous capabilities, which enable them to go beyond the simple representation of an architectural structure. They can provide information regarding the materials used and the realistic texture of the surfaces and also be interconnected with a data base for storing, managing and exploiting diverse information. Typical implementations of 3D modelling can be found in modern museums and educational foundations helping their visitors and students to communicate in a special way with the monument or site of interest as they can ‘walk’ through it or fly over it and thus examine it better, having always in mind its level of accuracy and likelihood. Researchers are also using 3D data acquisition to not only conserve excavation sites, but to reconstruct the actual excavation for real-time analyses (Levy 2013). 4. 3D VIRTUAL RECONSTRUCTION The term reconstruction implies the rebuilding of a monument to its state at a particular time moment of its past life, chosen for the representation. The term has similar meaning with the terms anastylosis and restoration, with the difference that the anastylosis is expected to use the authentic material, while for the restoration new material may be used, but both are implemented up to the point where assumptions about the original form of the monument are required. Nowadays archaeologists are extremely reluctant in actually reconstructing a monument for a number of reasons. Digital technologies have enabled the virtual reconstruction (Matini et al. 2008, Lentini 2009, Matini et al. 2009, PatayHorvath 2011, Vico & Vassalo 2008). This term implies that the representation takes place in a three dimensional space, which is usually called virtual environment and the final product is usually called a 3D virtual model. It is evident that 3D virtual reconstructions significantly support studies for the eventual real reconstruction © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 155-164 158 ANDREAS GEORGOPOULOS of the monument in the future. A virtual reconstruction would also enable the examination of various alternative solutions and help decisions for the suitable use of the salvaged members today. In case of monuments preserving most of their characteristics, or having been restored in the past, descriptive 3D models apply. In this case a geometric documentation with simple and suitable methods can generate 3D products which are good enough for visualization and can be obtained with varying degree of accuracy and detail. On the other hand, when we deal with objects that have few or practically no evidence of their past form and appearance, modelling is more complicated and needs hypotheses with different degrees of likelihood. There are many kinds of virtual reconstruction, mainly focussing on the verisimilitude of data used (Valle Melon et al. 2005, De Fuentes et al. 2010). This implies that data should be ranked according to their reliability and accuracy and given appropriate likelihood (Gkintzou et al. 2012, Kontogianni et al. 2013). All data collected are evaluated and used appropriately for the reconstruction. It is evident that, on one hand, the data collected do not all belong to the target period and, on the other, not all the data necessary to built up the model are available today. Therefore careful selection should always be done of the data corresponding to the period of study and complete them with suitable hypotheses. Usually the different data are illustrated with different colours, every colour representing different data source. Another way of representation is that the object parts are illustrated with different transparency level according to the reliability of original source. In this case important role plays the data date, their accuracy and their likelihood. The differentiation may also represent the knowledge about the original construction material. The application of the texture must be carried out carefully, because it is necessary to texture the virtual model of the monument correctly and make it look as real as possible. Very often, special decisions should be made during the progress of the 3D reconstruction, such as interpreting the geometry of the architectural elements like symmetry and the construction method of each element or section of the monument. Symmetry, which is a very important element in architecture, may also be represented and interpreted with the virtual reconstruction, while building the geometry of the architectural elements, like windows, doors, arcs etc. Another important element is the representation of the construction method of each element or section of the monument. Hence, the process of virtual reconstruction must be done very carefully in order to create the virtual model correctly both geometrically and photorealistically. 5. VIRTUAL RECONSTRUCTION Three examples of virtual reconstruction will be presented in order to illustrate their usefulness and great potential. All projects have been performed by the Laboratory of Photogrammetry of the National Technical University of Athens (NTUA). The first example is a standard virtual reconstruction of a monument which is ruined today. The second example is a case of a modern monument, which was restored on the basis of its virtual reconstruction. The final example is a case of a monument of which today nothing is salvaged, but its foundations and very few artifacts of the initial construction. They form a complete set of various cases of virtual reconstruction, which have been used for various purposes for the benefit of Cultural Heritage. 5.1 The Monastery of San Prudencio The Monastery of San Prudencio is located at the province of La Rioja in Spain. The three dimensional virtual restoration and reconstruction of the © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 155-164 3D VIRTUAL RECONSTRUCTION OF ARCHAEOLOGICAL MONUMENTS Figure 1: The Church of San Prudencio today church of the Monastery was performed within the framework of a larger project (Gkintzou et al. 2012). As the monument is almost completely ruined (Fig. 1), many additional sources were used for the reconstruction. It was decided to produce a 3D virtual model with surfaces in order to convey as much information as possible and represent the church as it probably was during the 14th-15th century according to historical sources that are available and other essential information selected in collaboration with archaeologists and architects. The restoration was also based on the detailed documentation of the current situation of the monument that conserves parts of the target phase. The documentation products were surveying measurements, Digital Surface Models (DSM), orthophotos, and laser scanner point clouds. The documentation of the current situation was needed for two reasons. Firstly because it provides information about the past of the monument and it is recommended for all related projects. Secondly, it includes existing elements, which will serve as the basis for the reconstruction phase. As far as the current situation is concerned digital images were acquired, which were used in order to generate Digital Surface Models (DSM) and orthophotos. In addition, surveying measurements and some laser scanner data were also collected. These images were taken from different angles and they aimed to record as completely as possible the relative position of the various remaining 159 structural elements and the materials on the wall. In general, they served as valuable reference, but they do not have any metric accuracy and they are not suitable for 3D modeling. It is evident that the data collected do not all belong to the target period and not all the data necessary to built up the model are available today. Therefore, we need to carefully select the data corresponding to the period of study and complete them with suitable hypotheses. It is imperative that both tasks must be done in collaboration of the archaeologists and architects. In this context a hierarchy of the data was developed, based on their reliability as far as their “correctness” is concerned (Table 1). Table 1: Reliability of data (1-10 decreasing) Sources Written Documents Reliability 2 Images 6 Drawings-Paintings 6 Surveying Measurements 3 Orthophotos 5 DSM 4 3D Line Drawings (Laser Scanner) 4 Archaeological Assumptions 1 Reliability measure, as depicted in Table 1, does not have anything to do with the accuracy of the surveying and photogrammetric measurements. Hence, the archaeological and architectural experts’ opinion and assumptions are considered as most reliable. Similarly, the rest of the various sources were evaluated by the interdisciplinary group. Written sources are high in the reliability scale, while drawings and paintings have a higher degree of subjectivity and are graded low in reliability. Surveying measurements and surface descriptions (DSM) as well as orthophotos are in the middle of the scale, as they do not describe only the specific era of reconstruction. The final 3D model was constructed by also taking into account the criterion of accuracy. In this case, there were three © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 155-164 160 ANDREAS GEORGOPOULOS sources giving information about an arc on the southern wall of the church. These sources were the DSM, the orthophotos and the 3D line drawings as they were extracted from the laser scanner data. As it was expected, they do not coincide as their accuracy differs. Consequently a second table was created (Table 2) representing the hierarchy of the data based on the criterion of accuracy. we will wait until the architectural and archaeological studies go further after deciding about them. Table 2: Accuracy of data (1-10 decreasing) Sources Written Documents Photos Drawings-Paintings Surveying Measurements Orthophotos DSM 3D Line Drawings (Laser Scanner) Archaeological Assumptions Geometric Accuracy 5 7 8 1 2 3 4 6 It is clear that 3D Virtual models representing reconstructed objects, that do not exist today, include elements with different levels of accuracy and likelihood. It has to be mentioned that the likelihood expresses the possibility of each element to exist during the period of reconstruction as it is presented at the model, while the accuracy describes the certainty related to the absolute and relative position of the elements. There are elements for example that can be represented with better accuracy than others, because they still exist but one cannot be so sure about their existence during the reconstruction period. In this case these two characteristics (likelihood and accuracy) of the representation coincided as the likelihood was depended on the kind of sources present for every element just like the accuracy level. If a more complete architecture and archaeological study were available this two characteristics may differ for some elements. All the data were examined critically in order to approximate the form and the structure of the church as well as the textures too. At the end, it was decided that the materials and textures did not need to be defined yet and, therefore, Figure 2: Virtual reconstruction with differentiation of data reliability In this way the final reconstruction (Fig. 2) reflects the various reliability grades, thus helping the potential user to comprehend it in a better way. This significantly differentiates this alternative virtual reconstruction approach, as it adds the reliability aspect. 5.2 The Zalongon Sculpture The monument of Zalongon is a huge complex of sculptures, 15m tall and 18m long, built in the 1950’s and located on top and at the edge of an 80m high cliff in north-western Greece. The monument commemorates the sacrifice of the Souli village women, who in 1803 preferred death from humiliation by the Ottoman conqueror. The restoration work that was carried out involved the cleaning of the sculpture’s surface, the extraction and replacement of large pieces that have suffered damages from harsh weather and were deteriorating rather quickly, and the restoration of parts that have been destroyed by frost. The data required to build high resolution 3D textured models include, 3D scans, geodetic measurements and a significant number of images. The detailed geometric documentation of the current situation of the monument included the production of 2D drawings, orthophotos and an accurate 3D model (Valanis et al. 2009). Engineers who are involved in restoration are greatly facilitated if they can © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 155-164 3D VIRTUAL RECONSTRUCTION OF ARCHAEOLOGICAL MONUMENTS interact with a 3D model and immediately obtain various kinds of information by measuring various distances, areas, volumes, by creating cross-sections, outlines or even by formulating and adding missing parts. However, in cases where the formulation and addition of 3D data is desired different methods and algorithms are required. This was also the case for the monument of Zalongon, where the upper parts of the two tallest figures were almost destroyed. Two main categories of data were extracted, namely the part of the surface that was healthy and would be retained and the broken part that was recorded only in order to help reconstruct what was missing. 161 original sculptor. The new plaster models were scanned with an XYZRGB SL2™ structured light scanner and the data acquired were registered with the 3D model. The final mesh was exported, appropriately scaled and in such a form to enable masonry experts to reproduce exactly the missing parts and to actually restore the monument (Fig. 3). 5.3 The Middle Stoa in the Athenian Agora The Ancient Athenian Agora is today one of the most important archaeological sites in Athens and is situated at the northern foot of the Acropolis hill. It was a Figure 4: The foundations of the Middle Stoa (a) (b) Figure 3: The 3D model before (a) and after (b) restoration Efforts were also made to completely restore the original surface virtually. However, in order to obtain a better result, another approach was preferred. The partially completed surfaces were used for the creation of analogue models of a scale 1:5 and an artist, a sculptor, was assigned with the task of completing the forms based on the existing model and old photographs and sketches made by the large open space to the south of Eridanos River and served as the administrative, philosophical, educational, social and economical centre of the town of Athens for many centuries. The Middle Stoa was an elongated building 147m by 17.5m, which ran east– west across the old square, dividing it into two unequal halves. This large building was constructed with Doric colonnades at both the north and south sides as well as an Ionic colonnade along the middle. The original steps and three columns remain in situ at its eastern end; to the west, only the heavy foundations of reddish conglomerate survive. The Middle Stoa was built between ca. 180 and 140 B.C. and it was continuously used even during of the Roman era (www.agathe.gr). Foreign architects were responsible for its construction; hence it presents particular design and construction elements not usual for that time in the area. Today only the foundations of this majestic building and some individual parts of it are visible in the site (Fig. 4). © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 155-164 162 ANDREAS GEORGOPOULOS For the virtual reconstruction several different data were available. Artifacts from the initial construction in the museum, drawings from scholars who had studied thoroughly the monument (Travlos 1971, Muller-Wiener, 1995), survey measurements of the foundations which are visible today and artists’ reproductions of pertinent descriptions from travellers of the past. All data were evaluated (Table 3) for their reliability and accuracy before usage for the final virtual reconstruction (Fig. 5) Table 3: Data evaluation Characteristics Data Source 3D Model Archit. plans Other plans Images Literature Assumptions Year 2010, 2012 1963, 1966 Varies Varies Varies - Accuracy Likelihood 1 1 2 2 3 5 4 6 3 5 4 6 Figure 5: The reconstructed Middle Stoa This final product reconstructs a building that does not exist today. The visitor may only see the foundations of the building, which at the time of its peak (2nd c. BC) were buried in the ground. Consequently the virtual reconstruction is a combination of existing detailed architectural drawings, of sketches, descriptions, digitization of real artefacts and other minor sources of information. Of utmost importance were the discussions and suggestions of scientists who have studied the monument from an historical and archaeological point of view, proving once again that a reconstruction is a multi disciplinary process. 6. CONCLUDING REMARKS The final products are virtual reconstructions of buildings that do not exist today. Consequently virtual reconstructions are combinations of existing detailed architectural drawings, of sketches, descriptions, digitization of real artefacts and other minor sources of information. Of utmost importance are the discussions and suggestions of scientists who have studied the monuments from an historical and archaeological point of view, proving once again that reconstructions are a multi disciplinary process. Virtual reconstructions on the other hand support many other disciplines involved in cultural heritage. They help architects in their work for monuments especially in cases of restoration, anastylosis etc. Archaeologists and Conservationists have a very good tool at their disposal for their studies. Many applications can be generated from a virtual reconstruction like virtual video tours of the monument for educational and other purposes for use by schools, museums and other organizations, for incorporation into a geographical information system (GIS) for archaeological sites, for the design of virtual museums and the creation of numerous applications for mobile devices (e.g. mobile phones, tablets etc). Virtual reconstructions have the undeniable advantage that they do not harm the existing architectural elements and, most importantly, they may be reproduced in many different ways depicting each time a different solution to the inevitable questions that arise during a reconstruction process. Thus they could be a very powerful tool for in depth studies of our Cultural heritage. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 155-164 3D VIRTUAL RECONSTRUCTION OF ARCHAEOLOGICAL MONUMENTS 163 REFERENCES CIPA Recordim http://cipa.icomos.org/index. php?id=43 (last accessed 28.11.2013) De Fuentes F. A., Valle Melon J.M., Rodriguez Miranda A. 2010. Model of sources: a proposal for the hierarchy, merging strategy and representation of the information sources in virtual models of historical buildings. CAA conference Proceedings, "Computer Applications and Quantitative Methods in Archaeology", Granada. Gkintzou, Ch., Georgopoulos, A., Valle Melón, J.M., Rodríguez Miranda, Á., 2012. Virtual Reconstruction of the ancient state of a ruined Church. Project Paper in “Lecture Notes in Computer Science (LNCS)”, Springer Verlag. 4th International EuroMediterranean Conference (EUROMED), 29/10 – 03/11 2012 Limassol Cyprus. Kontogianni, G., Georgopoulos, A., Saraga, N., Alexandraki, E. and Tsogka, K., 2013. 3D Virtual Reconstruction of the Middle Stoa in the Athens Ancient Agora, Int. Archives of Photogrammetry, Remote Sensing and Spatial Information Science, XL-5/W1, 125-131, doi:10.5194/isprsarchives-XL-5-W1-125-2013, 2013. http://www.int-arch-photogramm-remote-sens-spatial-inf-sci.net/XL-5W1/125/2013/isprsarchives-XL-5-W1-125-2013.pdf Lentini D., 2009. The funeral area in “Ponte Della Lama Canosa” (III-VI century) an hypothesis of 3D historical- monumental reconstruction, Proceedings of 3DArch, Trento, Italy, February 25-28. Levy, T.E. 2013. Cyber-Archaeology and World Cultural Heritage: Insights from the Holy Land. Bulletin of the American Academy of Arts & Sciences LXVI:26-33. Matini M.R., Einifar A., Kitamoto A., Ono K., 2009. Digital reconstruction based on analytic interpretation of relics: case study of Bam citadel. XXII International Symposium of CIPA, Kyoto, Japan. October 11-15. Muller-Wiener E., 1995. The architecture in Ancient Greece. University Studio Press, ISBN 9601204849, pp.247 Thessaloniki (in Greek). Patay-Horvath. 2011. The complete virtual 3D reconstruction of the east pediment of the temple of Zeus at Olympia. Proceedings of 3DArch, Trento, Italy, March 2-4. Matini, M.R., Andaroodi, E., Kitamoto, A., Ono, K., 2008. Development of CAD-based 3D drawing as a basic resource for digital reconstruction of Bam’s Citadel (UNESCO World Heritage in danger), EuroMed 2008: Digital Heritage – Proceedings of the 14th International Conference on Virtual Systems and Multimedia, M. Ioannides, A. Addison, A. Georgopoulos, L. Kalisperis (Full Papers), pp. 51-58. Rodríguez Miranda, Á., Valle Melón, J. M., Martínez Montiel, J. M., 2008. 3D line drawing from point clouds using chromatic stereo and shading. VSMM 2008, Digital Heritage - Proceedings of the 14th International Conference on Virtual Systems and Multimedia. 20-26 November 2008. Limassol (Cyprus). Travlos J., 1971. Pictorial dictionary of Ancient Athens, Thames & Hudson, ISBN 0500050120, pp. 590, London. Valanis, A., Georgopoulos, A., Tapinaki, S., Ioannidis, Ch., 2009. High Resolution Textured Models for Engineering Applications. Proceedings XXII CIPA Symposium, October 11-15, 2009, Kyoto, Japan, http://cipa.icomos.org/text%20files/KYOTO/164.pdf Valle Melon J.M., Lopetegi Galarraga A., Rodriguez Miranda A. 2005. Problems when generating virtual models representing real objects: Hondarribia walls. Proceedings of Virtual Retrospect, Biarritz, France. November 2-10. Vico, L., Vassallo, V., 2008, The Reconstruction of the Archaeological Landscape through Virtual Reality Applications: a Discussion about Methodology, EuroMed 2008: Digital Heritage – Proceedings of the 14th International Conference on Virtual © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 155-164 164 ANDREAS GEORGOPOULOS Systems and Multimedia, M. Ioannides, A. Addison, A. Georgopoulos, L. Kalisperis (Full Papers), pp. 397-403. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 155-164 Mediterranean Archaeology and Archaeometry, Vol. 14, No 4, pp. 165-174 Copyright © 2014 MAA Printed in Greece. All rights reserved. THE DEVELOPMENT OF VOCABULARIES OF HISTORICAL PERIOD NAMES FROM WEB ACQUIRED CORPORA7 Maria S. Mouroutsou*1, Stella Markantonatou2 and Vasilis Papavasiliou2 1National and Kapodistrian University of Athens, School of Philosophy, Dept. of Philology Panepistimiopolis, 15784 Ilissia, Greece 2ILSP/ “Athena”RIC, Artemidos 6 & Epidavrou GR- 151 25 Maroussi, Greece Received: 15/12/2013 Accepted: 08/05/2014 Corresponding author: Maria S. Mouroutsou (msmourou@gmail.com) ABSTRACT Periodization is a universal and very popular system of organizing History (Petras, et al., 2006) by arbitrary dividing time into periods such as “Δικτατορία” (dictatorship) in a way that is specific to places and communities. Structured collections of time period names and timelines are considered very useful in cultural content documentation and temporal information extraction. However, to the best of our knowledge, this is the first report on the systematic collection of period names of Greek History. New period names are constantly created or left out of use. Aiming to capture this combination of dispersed specificity and constant evolution, we used the Focused Monolingual Crawler (FMC) (Mastropavlos, et al., 2011) and an initial list of 25 “seed-terms” to develop corpora dense in period names with Web retrieved documents. Period names were manually retrieved from the accumulated corpora and were annotated for a set of features, including allomorphs that occurred in the collected corpora and whether the term denoted a fact or a time period or something else as well as for persons, places and other period names related with the term. The linguistic environments where the terms occurred were identified and some of them were fed to the (FMC) as new “seed-terms”. This cycle was repeated for three times and yielded 78 period names with an average of 16 paradigms per term and a corpus consisting of 3020 valid XML documents. Some first observations on the strategies employed by Greek communities to coin time period names are reported. KEYWORDS: periodization, time period name, Focused Monolingual Crawler, unstructured Web data 166 MARIA S. MOUROUTSOU et al 1. INTRODUCTION Periodization is a system of organizing historical information by dividing time into periods. Time period names, a type of named entity in fact, often denote more than simply calendar dates; they implicate a subject, time and place (ISO/CD21127). For example, the Greek period name “χρυσούς αιών” (golden age) encapsulates a place (Athens), time (fifth century), and subject (a flowering of arts and culture, and the height of democracy). In CIDOC CRM (Doerr 2003; Crofts et al., 2004) the basic notion of a period is defined as follows: “This class comprises sets of coherent phenomena or cultural manifestations bounded in time and space. It is the social or physical coherence of these phenomena that identify a…period and not the associated spatio-temporal bounds. These bounds are a mere approximation of the actual process of growth, spread and retreat. Consequently, different periods can overlap and coexist in time and space, such as when a nomadic culture exists in the same area as a sedentary culture.” (Crofts et al., 2004). In Feinberg et al, (2003) time periods are described more or less as gazetteers that match place names to coordinates. Just like a gazetteer, a time period directory could match time period terms to date ranges, location and other information that characterizes the period. To these descriptions we would add that time period names are constantly developed and abandoned; the phenomenon is of a dynamic nature. In the light of the above descriptions of the notion ‘time period’, it is only natural to say that attempts at periodization are never neat: one period flows into another or the same name may be used for periods that occur at different times in different regions, for instance, “εμφύλιος” (civil war) is a time period that determines different times from the ancient history of Greece to our days. However, time period names are used widely in both the everyday language and the scientific jargon and are very useful in documentation. Furthermore, their study could reveal how human communities, here the Greek speaking communities, coin period names, for example which events lend their names and which linguistic means are employed to coin a time period name. To the best of our knowledge, there is no structured collection of time period names available about Modern Greek History; indicatively, in the relevant lemma in the Wikipedia such information is extremely lean. Furthermore, there is no study concerning the linguistic characteristics of these terms. So, the first step is to develop a corpus of time period names. The second step is their organization into ontologies and timelines. In this paper we report on the first step. We report on the development of vocabularies of historical period names from Web acquired specific corpora. Web data would cater for the sparcity of Greek corpora dense in time period names and for the dynamic nature of these terms. We focused on period names of the 19th and 20th century of Greek History. For data collection from the Web we used the ILSP - Focused Monolingual Crawler (FMC) (Mastropavlos, et al., 2011). This paper is organized as follows: Section 2 briefly presents related work. Section 3 describes the methodology we developed to collect Web corpora and identify the time period names. Section 4 introduces some observations concerning the strategy Greek communities use to coin new time period names. Conclusions are given in Section 5 and plans for future work in Section 6. 2. RELATED WORK Time period names have only recently attracted the ICT research community because Semantic Web has increased needs in standardized documentation and linked data. Structured collections, for instance ontologies, of time period names are sparse however (Berman, 2011). Feinberg et al., (2003) pioneered work on time period names. They treat them as gaz- © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 165-174 THE DEVELOPMENT OF THESAURI OF HISTORICAL PERIOD NAMES etteers and present a Content Schema that specifies the information required to define a period name. Furthermore, drawing on data from the University of California MELVYL catalog and several timelines available on the Web, they discuss the issue of distinguishing between time period names and event names. Although duration is a semantic criterion that favors the time period reading of a term, it is not always conclusive. For instance, they notice that Kennedy assassination is used as a micro-era period name despite the fact that it clearly refers to an instantaneous event. The authors propose that a set of criteria are used including the role of the name in context and duration of the time interval denoted. Petras et al. (2006) developed a prototype Time Period Directory with time period names and events extracted from the Library of Congress subject headings. Drawing on Feinberg et al., (2003), they too argue that a time period directory could work as a place name gazetteer does because (1) as locations are referred to by place names similarly spans of time are commonly referred to by period names, such as “Napoleonic wars” and (2) time periods have a geographical aspect as gazetteer entries have a period aspect (Buckland, et al., 2004). They developed a Time Period Directory with 2,000 entries derived from the Library of Congress Subject Headings. Drawing on the Feinberg et al. (2003) Content Schema, they proposed a Content Standard and the relevant XML Schema that has been adopted by the ECAI (http://ecai.org/). The DDBC Time Authority Database (http://authority.ddbc.edu.tw/docs/open _content/) is one of a group of Authority Databases provided by Dharma Drum Buddhist College (DDBC). It contains detailed Chinese calendar data from the beginning of the Qin dynasty to the current day. The major purpose of DDBC is to provide complete Chinese calendar information that can be used by external services and applications. Time period names 167 play an important role in the organisation of the DDBC data. Time period names have been considered as crucial parts of next generation gazetteers (Berman, 2011) that will link together Place Names (Place Name Authorities, Historical GIS and geonames) with Chronologies (Administration Periods, Timelines of events and Time Period Names Index) and Entity Definitions. 3. METHODOLOGY FOR COLLECTING (GREEK) TIME PERIOD NAMES We report on the development of collections of Greek time period names. This work consists of two parts: (1) collection of Web corpora dense in time period names (2) multidimensional annotation of terms. Our work differs from existing work on period names in that there were no structured resources, such as the Library of Congress archives or substantial timelines, to refer to. Instead, we had to resort to fully unstructured resources and in fact, discover ways of collecting the right ones. Furthermore, given that the contexts where the terms occurred were of a general nature, we had to disambiguate among more term readings than simply those of a period name and an event. We confined ourselves to the 19th and the th 20 century of Greek history (1821: the successful Greek Independence Revolution against the Turkish Occupation) in order to simplify things. To develop a rich collection of period names from the Greek History of the 19th and the 20th century, a specific domain corpus with different types of text should be used. The need for a large variety of texts is intensified by the fact that period names should be treated as a dynamic phenomenon. The available Greek corpora are of small or medium size and certainly not dense in such terms. The existing general purpose Greek corpora, Hellenic National Corpus being the standard example (Gavrilidou, 2002), as well as Corpus Greek Texts (Goutsos, 2010) have not been built for use in specialized domains, such as history, medicine or education. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 165-174 168 MARIA S. MOUROUTSOU et al Fortunately, tools for easily developing corpora rich in the respective material of interest are available nowadays. We used a revised version of the Focused Monolingual Crawler (FMC) (Mastropavlos, et al. 2011) developed by the Institute for Language and Speech Processing (ILSP/ “Athena” RIC). The collected corpora were processed manually to extract candidate time period names and their contexts. Then, a battery of linguistic criteria was used to distinguish between names that denoted time periods and names that denoted something else, most often events. gual domain-specific data is illustrated in Figure 1. 3.2 Experiments with the FMC Three experiments were conducted with FMC. Greek was used for all the three experiments. Greek is not an under-resourced language, as the basic information technology is available and has a relatively substantial presence in the Web (Berment, 2004). On the other hand, there is a clear need for large Greek corpora and electronic lexica and this fact underlines the importance of FMC for Greek. 3.1 Focused Monolingual Crawler (FMC) FMC is a system that explores the Web and downloads pages with text related to a specific domain. Also, it includes components for all the tasks required to acquire domain-specific corpora from the Web. Due to its modular architecture, each of these components can be easily substituted with alternatives of the same functionality. Furthermore, the system is available as an open-source Java project (http://nlp.ilsp.gr/redmine/projects/ilspfc/). Comprehensive documentation on how to get setup and run it, is available (http://nlp.ilsp.gr/redmine/projects/ilspfc/wiki). Given a narrow domain defined by a certain topic and a language, FMC is fed with two input datasets: (i) a list of topic definition (multi-)word terms and (ii) a list of topic related URLs (Skadina et al., 2012). The user can configure FMC in a variety of ways, for instance set file types to download, domain filtering options, selfterminating conditions, crawling politeness parameters, etc”. The crawler first visits the web pages that are initially provided by the user and fetches related ones. Next, it classifies fetched pages as relevant to the targeted domain and extracts links from fetched web pages. Finally, it adds them to the list of pages to be visited and repeats this cycle. A typical workflow for acquiring monolin- Figure 1 A typical workflow of FMC. In all the three experiments the input to the FMC consisted of a seed URL list that initialized the crawler’s frontier. In our case, a possible initial web page would be the Wikipedia page for “Τουρκοκρατία” (Turkish Occupation). As explained in Section 3.1, the input also included the domain definition that consisted of terms describing the domain. We set off with a TermList of twenty five terms. Figure 2 shows a subset of these terms. The terms were collected from the High School History textbooks. Our aim was to create a first sample, a list just to feed the crawler. In this first experiment only a dozen of URLs retrieved from search engines were given as an input. Mostly, they were Web pages of museums, libraries and wikis. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 165-174 THE DEVELOPMENT OF THESAURI OF HISTORICAL PERIOD NAMES 169 In all the three experiments, we set the time parameter of the crawler to a single hour of work. We arrived to this decision because we wanted the size of our corpus to be manageable for a manual retrieval of period names. way we identified several syntactic environments and enriched our TermList for the experiments that followed. TermList_Exp1 Β’ Παγκόσμιος Πόλεμος-2nd World War Ελληνική επανάσταση-Greek revolution TermList_Exp2 Τουρκοκρατία-Turkish Occupation Οθωνική περίοδος-Otto period Ανατολικό Ζήτημα-Eastern Question Περίοδος ανεξαρτησίας-Independence period Περίοδος αντιβασιλείας-Regency period Δικτατορία του Μεταξά Metaxas’ Dictatorship Περίοδος της Αντίστασης-Resistance Period Στα χρόνια Κατά την περίοδο Α’ Βαλκανικοί Πόλεμοι-1st Balcan Wars Την εποχή 20ος αιώνας-20th century Ελληνοϊταλικός Πόλεμος-Greek-Italian War Μικρασιατική Καταστροφή Asia Minor Destruction Εμφύλιος Πόλεμος-Civil War Κατοχή-German Occupation Δικτατορία του Παπαδόπουλου Papadopoulos’ Dictatorship Μεταπολίτευση (period immediately after the 1967-1973 Junta) Figure 2 TermList – Experiment 1. The output of the first experiment contained almost 500 URLs. With an advanced search we found out that the result would be better if our TermList was enriched with more period names. Apart from period names, in the subsequent two experiments we used key-words that were not historical periods, such as the name of the protagonist of a fight, a war or a military movement as well as names of locations. Moreover, we added words that were found in the contexts where historical names occurred. The aim was to enhance our searching tools and obtain more results relevant to the periods in question. In order to identify such words, we studied the contexts where the term “Τουρκοκρατία” (Turkish Occupation) occurred. Since the period is dominant in Modern Greek history, we knew that there would be a big amount of documentation for it. In that Τα γεγονότα της Κατά τη διάρκεια Επί πρωθυπουργίας Figure 3 TermList – Experiment 2. Last, we noticed that certain words functioned as heads of period names such as the word “πόλεμος” (war) or “εποχή” (era) or even “περίοδος” (period). Τhose words were also included in our term list. Furthermore, in the next two experiments we enriched the URLs list with web pages that would use more colloquial language and would capture the dynamic nature of period names, such as newspapers, blogs and forums. Fig. 3 shows the results of the three experiments: Parameters Ex1 Ex2 Ex3 TermList 25 75 180 Language Gr Gr Gr URLlist 12 25 50 MaxTime 1h 1h 1h 911 1622 Results XML/URL 487 Figure 4 The input and the output of the three experiments. 3.3 Annotation We annotated our corpus manually. We retained a 50-word context before and after © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 165-174 170 MARIA S. MOUROUTSOU et al the terms of interest. Apart from this information, terms were annotated for the following features: syntactic structure (for instance Adjective+Noun), syntactic role in the text (subject, object etc), allomorphs of the term in the collected corpora, close context that helped disambiguating the reading of the term between that of a time period and that of an event, whether the term denoted a fact or a time period, whether the term denoted a specified time period or part of a specified time period, persons, places and other period names related with the term, the actual dates related with the term and the URL where the term was retrieved from. We used contextual information to disambiguate between a time period reading and other readings, the event reading being the dominant but not the only alternative used. As a working example we will use the term Εθνική Αντίσταση (National Resistance) that refers to the Greek resistance during the German Occupation (19411944). The term returned 25 examples, of which 8 were classified as period names. We first set apart metonymic usages such as ‘agent’ (1) and ‘moral imperative’ (2). (1) H Εθνική Αντίσταση κέρδισε την πρώτη μάχη της σε ανοιχτή αναμέτρηση με τις δυνάμεις της φασιστικής βίας. ‘National Resistance wined its first battle in an open struggle against the powers of fascist violence’. (2) Αντίθετα, μετά την Εθνική Αντίσταση του ΕΑΜ, το πρόταγμα για Λαοκρατία κρατά συμπαγές το ΚΚΕ... ‘On the contrary, after the National Resistance, the ethical imperative Rule-Of– the-People pulls the Greek Communist Party together’ … Next, we classified as event names terms that occurred in the following contexts: –When the term was used to explain the word ‘event’ -Contexts where a set of actions were denoted. For instance, in (3), contrary to an event, a time period just occurs and can not be ‘organised’. A similar context is ‘X took an active role in <TERM>’ - Lists of events that included the term in question. (3) Στην αρχή του πολέμου ο κόσμος έμεινε άφωνος... αλλά στη συνέχεια οργανώθηκε η Εθνική Αντίσταση ‘When the war started, people were left speechless… but then the National Resistance was organized’. We classified as time period names terms that occurred in the contexts listed below: - Titles (ambiguous between an event and a time period reading) - The term offered the time contour for the event that was described (4) - Contexts where a ‘set of coherent …cultural manifestations’ reading (Crofts et al., 2004) was allowed (5). (4) Μετά από όλες τις ταλαιπωρίες που πέρασε πολεμώντας στην Μικρά Ασία,..., κατόπιν στο Αλβανικό και αργότερα στην Εθνική Αντίσταση όπoυ ήταν Διοικητής στο 52 Σύνταγμα του ΕΛΑΣ στο Λιανοκλάδι... ‘After all the suffering he went through fighting in Asia Minor, …, then in the Albanian War and later during the National Resistance when he was Commander of the 52th Regiment of ELAS at Lianocladi…’ (5) …διοργανώνει, σε συνεργασία με το Δήμο Ηλιούπολης, εκδήλωση με θέμα: «Η Ελληνίδα Γυναίκα στην Εθνική Αντίσταση και η σύγχρονη κοινωνική πραγματικότητα. ‘…organizes, in cooperation with the Ilioupoli Municipality, a discussion with the subject “The Greek Woman at the National Resistance and the modern social situation’. As is the rule, term sense disambiguation is not a clear-cut job. We plan to include some inter-annotation experiments in our future work. We started our study with only 19 period names and no contexts of usage. We now have a list of 78 period names, each one exemplified with an average number of 16 contexts of usage. Given that we achieved these results with only 3 hours of Web corpora collection with FMC, we conclude that our method can be used © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 165-174 THE DEVELOPMENT OF THESAURI OF HISTORICAL PERIOD NAMES successfully to develop corpora dense in domain specific information that evolves dynamically. Certainly, identification of time period names in the accumulated corpora is a laborious task that is performed manually given the present state-of-the-art in NLP in terms of recall and precision. However, the collection of appropriate corpora is actually the hardest and least attractive part of the work. Our method reduces to a minimum the effort for appropriate corpora collection. In addition, the annotated corpora that have been developed at the course of this work form a recourse that could be used as a training/test set for developing a module for automatic identification of period names. 4. CONSTRAINTS ON THE CREATION OF TIME PERIOD NAMES A first picture of the semantic constraints that control the process of coining time period names has emerged from our work. And in fact, as native speakers of Greek, we were surprised to find out that the Greek communities do not coin period names out of names of treaties, military movements and battles. Figure 4 offers an idea of the overall picture. Period Names Events Καποδιστριακή Περίοδος The period of Capodistrias Παλινόρθωση When the Greek King resumed his reign Χρόνια της Κρητικής Πολιτείας The years of the Cretan State Κίνημα στο Γουδί (military) Movement at Goudi Συνθήκη των Σεβρών Treaty of Sevres Μικρασιατική Καταστροφή Asia Minor Destruction Εθνικός Διχασμός National Division Δεκεμβριανά The events of December Κίνημα του Ναυτικού Movement of the Navy Πραξικόπημα Coup Εμφύλιος Civil (War) Είσοδος της Ελλάδας στην Ε.Ε. Admission of Greece to the EU 171 Μοναστηριακό ζήτημα Monastery Issue Κρητική Επανάσταση Cretan Revolution Figure 5 Period Names – Facts/Other Readings. The explanation that treaty signing, military movements and battles are events of a rather short duration may be put forward for this tendency of Greek. However, strong counterexamples to this explanation exist. For instance, we were very surprised to find out that “Μικρασιατική Καταστροφή” (Asia Minor Destruction) that describes an extended period (at least the biggest part of the year 1922) was never used with a time period reading in our data. 5. CONCLUSIONS We have described a method for collecting a substantial number of period names from unstructured Web data. The method involves techniques for collecting specific domain corpora dense in the desired terms by using the Focused Monolingual Crawler of ILSP. The method also comprises the identification of linguistic contexts where period names occur. The method is useful to languages that have a presence in the Web but have no corpora appropriate for time period name retrieval. In addition, the method may be useful to languages with richer resources because time period names are a dynamic phenomenon that can be better described with corpora that reflect the current usage of language. 6. FUTURE WORK In the immediate future we plan to take the following steps: -Verify our sense disambiguation results with targeted inter-annotation experiments -Use the Content Standard developed by Petras et al. (2006) and perhaps extend it with necessary information of a linguistic © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 165-174 172 MARIA S. MOUROUTSOU et al nature in order to encode as gazetteers the time period names we have collected -Develop an ontology of Greek period names that will be useful to ICT applications and lexicography - Enhance/ modify existing modules (e.g. TimeEL proposed by Prokopidis et al 2009) for automatic recognition of temporal expressions in Greek texts -Develop timelines for the Wikipedia and for educational purposes (visualization of data). As Greece is a country consisting of very many small places, each one with its own varied and turbulent history, such timelines will be of immense help not only to students and teachers but to the layman who is interested in history—let alone the researchers. 1 An excellent example of the educational usages of such timelines is the Greek term ‘Τουρκοκρατία’ (Turkish Occupation). Here, we describe the complications of the case of only one Greek town, namely the town of Naplion that is situated in EasternCentral Peloponnese. There are two periods with the same name for this town. The first one was between 1540 and 1687 and the second one was from 1715 to 1822. The situation gets very complex if all the areas where the term applies (they are many more than the ones included in the today Greek territory) are considered as there are areas that were under the Turks for 16 years only and areas that were under the Ottoman regime for five centuries. 1 © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 165-174 THE DEVELOPMENT OF THESAURI OF HISTORICAL PERIOD NAMES 173 REFERENCES Berman, M. (2011). Extending Gazetteers with Time and Entity Relationships. Historical Gazetteer Elements: Temporal Frameworks. Track on Historical Gazetteers Part of the Symposium on Space-Time Integration in Geography and GIScience co-sponsored by Harvard Univeristy’s Center for Geographic Analysis and the AAG, Wednesday-Friday, April 13-15, AAG 2011, Seattle, WA Buckland, M. and Lancaster, L. (2004) Combining Time, Place, and Topic: The Electronic Cultural Atlas Initiative, D–Lib Magazine, volume 10, number 5 (May), at http://www.dlib.org/dlib/may04/buckland/05buckland.html, accessed 2 June 2006. Crofts, N., Doerr, M., Gill, T., Stead, S. and Stiff, M. (2004) Definition of the CIDOC Conceptual Reference Model (version 4.0). DDBC Time Authority Database (http://authority.ddbc.edu.tw/docs/open_content/) Doerr, M., Kritsotaki, A. and Stead, St. (2003) Thesauri of Historical Periods – A Proposal for Standardization, (http://www.cidoc-crm.org/). Feinberg, M., Mostern, R., Stone, S. and Buckland, M. (2003) Application of Geographical Gazetteer Standards to Named Time Periods. Technical Report, Electronic Cultural Atlas Initiative, Berkeley. Gavrilidou, M., (2002) The Hellenic National Corpus on-line, Revue Belge de Philologie et Historie 80, pp. 1003-1015 Goutsos, D., (2010) The Corpus of Greek Texts: a reference corpus for Modern Greek, Corpora. Vol 5, pp. 29-44. Harvard University. Chinese Historical GIS Project. Available at <http://www.fas.harvard.edu/~chgis/>. ISO/CD 21127 (2002) Information and documentation – A reference ontology for the interchange of cultural heritage information Mastropavlos, N. and Papavassiliou, V. (2011) Automatic Acquisition of Bilingual Language Resources. In Proceedings of the 10th International Conference of Greek Linguistics, Komotini, Greece. Petras, V., Meiske, M., Larson, R., Zernecke, J., Carl, K. and Buckland, M. (2005) Leveraging Library of Congress Subject Headings to improve Search for Events – A Time Period Directory. Petras, V., Larson, R. and Buckland, M. (2006) Time Period Directories: a Metadata Infrastructure for Placing Events in Temporal and Geographic Context. Joint Conference on Digital Libraries, Chapel Hill, NC, USA. Prokopidis, P., Desipri, E., Papageorgiou, H. and Markopoulos, G. (2009) TimeEL: Recognition of Temporal Expressions in Greek texts. In Proceedings of the 9th International Conference of Greek Linguistics, Chicago, Illinois, USA. Skadiņa, I., Aker, A., Μαστροπαύλος, Ν., Su, F., Tufis, D., Mateja, V. et al. (2012). Collecting and Using Comparable Corpora for Statistical Machine Translation. In On-Line Proceedings of the LREC2012 Conference on Language Resources and Evaluation, pages 438-445. Istanbul, Turkey. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 165-174 174 MARIA S. MOUROUTSOU et al Support for the Learner: Time Periods, at http://ecai.org/imls2004/timeperiods.html, accessed 2 June 2006. Wikipedia. List of Themed Timelines. 2004. Available at <http://en.wikipedia.org/wiki/List_of_themed_timelines>. © University of the Aegean, 2014, Mediterranean Archaeology & Archaeometry, 14, 4 (2014) 165-174 Subscription Form YES, I would like to apply for an annual subscription to MEDITERRANEAN ARCHAEOLOGY AND ARCHAEOMETRY Name: ............................................................................................................................................... E‐mail address: ................................................................................................................................ Position: ............................................................................................................................................ Organization: ................................................................................................................................... Department: ..................................................................................................................................... Address: ........................................................................................................................................... Post/Zip Code: ................................................ Country: ............................................................... Tel.: ................................................................... Fax.: ....................................................................... Subscription rates: LIBRARIES: 100 EURO per volume (= 2 issues per year), INDIVIDUALS: 80 EURO per volume, STUDENTS: 50 EURO per volume. Back issues: as above according to category. Payment: From Greece: deposit the amount to ALPHA BANK 601002001000131 and send a fax copy of the receipt to: 210 6492499 c/o A. Vogiatzi. From Abroad: send bank cheque to c/o Prof. I. Liritzis, Aegean University Property Management Corp. In any case, notify us by e‐mail (maa_journal@rhodes.aegean.gr) when payment is made. For any inquiry contact address: MEDITERRANEAN ARCHAEOLOGY AND ARCHAEOMETRY, Laboratory of Archaeometry, c/o Prof. I. Liritzis Department of Mediterranean Studies, University of the Aegean, 1 Demokratias Av., Rhodes 851 00, Greece Tel: 0030 22410 99385‐6, Fax: 0030 22410 99320