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