ISPRS Ann. Photogramm. Remote Sens. Spatial Inf. Sci., IV-2/W6, 2019
This paper presents a method for the classification of images of silk fabrics with the aim to pre... more This paper presents a method for the classification of images of silk fabrics with the aim to predict properties such as the place and time of origin and the production technique. The proposed method was developed in the context of the EU project SILKNOW (http://silknow.eu/). In the context of classification, we address the problem of limited as well as not fully labelled data and investigate the connection between the distinct variables. A pre-trained Convolutional Neural Network (CNN) is used for the feature extraction and a classification network realizing Multi-task learning (MTL) is trained based on these features. The training procedure is adapted to enable the consideration of images that do not have a label for all tasks. Additionally, MTL with fully labeled training data is investigated for the classification of silk fabrics. The impact of both MTL approaches is compared to single-task learning based on two different class structures. We achieve overall accuracies of 92-95% and average F1-scores of 88-90% in our best experiments.
ISPRS Annals of Photogrammetry, Remote Sensing and Spatial Information Sciences
Supervised machine learning needs high quality, densely sampled and labelled training data. Trans... more Supervised machine learning needs high quality, densely sampled and labelled training data. Transfer learning (TL) techniques have been devised to reduce this dependency by adapting classifiers trained on different, but related, (source) training data to new (target) data sets. A problem in TL is how to quantify the relatedness of a source quickly and robustly, because transferring knowledge from unrelated data can degrade the performance of a classifier. In this paper, we propose a method that can select a nearly optimal source from a large number of candidate sources. This operation depends only on the marginal probability distributions of the data, thus allowing the use of the often abundant unlabelled data. We extend this method to multi-source selection by optimizing a weighted combination of sources. The source weights are computed using a very fast boosting-like optimization scheme. The run-time complexity of our method scales linearly in regard to the number of candidate sou...
ISPRS - International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences
We propose a new approach for the automatic detection of network structures in raster data. The m... more We propose a new approach for the automatic detection of network structures in raster data. The model for the network structure is represented by a graph whose nodes and edges correspond to junction-points and to connecting line segments, respectively; nodes and edges are further described by certain parameters. We embed this model in the probabilistic framework of marked point processes and determine the most probable configuration of objects by stochastic sampling. That is, different graph configurations are constructed randomly by modifying the graph entity parameters, by adding and removing nodes and edges to/ from the current graph configuration. Each configuration is then evaluated based on the probabilities of the changes and an energy function describing the conformity with a predefined model. By using the Reversible Jump Markov Chain Monte Carlo sampler, a global optimum of the energy function is determined. We apply our method to the detection of river and tidal channel ne...
ISPRS Annals of Photogrammetry, Remote Sensing and Spatial Information Sciences, 2015
The fully automated analysis of 3D point clouds is of great importance in photogrammetry, remote ... more The fully automated analysis of 3D point clouds is of great importance in photogrammetry, remote sensing and computer vision. For reliably extracting objects such as buildings, road inventory or vegetation, many approaches rely on the results of a point cloud classification, where each 3D point is assigned a respective semantic class label. Such an assignment, in turn, typically involves statistical methods for feature extraction and machine learning. Whereas the different components in the processing workflow have extensively, but separately been investigated in recent years, the respective connection by sharing the results of crucial tasks across all components has not yet been addressed. This connection not only encapsulates the interrelated issues of neighborhood selection and feature extraction, but also the issue of how to involve spatial context in the classification step. In this paper, we present a novel and generic approach for 3D scene analysis which relies on…
ISPRS Annals of Photogrammetry, Remote Sensing and Spatial Information Sciences, 2015
Tracking-by-detection is a widely used practice in recent tracking systems. These usually rely on... more Tracking-by-detection is a widely used practice in recent tracking systems. These usually rely on independent single frame detections that are handled as observations in a recursive estimation framework. If these observations are imprecise the generated trajectory is prone to be updated towards a wrong position. In contrary to existing methods our novel approach uses a Dynamic Bayes Network in which the state vector of a recursive Bayes filter, as well as the location of the tracked object in the image are modelled as unknowns. These unknowns are estimated in a probabilistic framework taking into account a dynamic model, and a state-of-the-art pedestrian detector and classifier. The classifier is based on the Random Forest-algorithm and is capable of being trained incrementally so that new training samples can be incorporated at runtime. This allows the classifier to adapt to the changing appearance of a target and to unlearn outdated features. The approach is evaluated on a publicl...
ISPRS - International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, 2015
In this paper we address the problem of classification of remote sensing images in the framework ... more In this paper we address the problem of classification of remote sensing images in the framework of transfer learning with a focus on domain adaptation. The main novel contribution is a method for transductive transfer learning in remote sensing on the basis of logistic regression. Logistic regression is a discriminative probabilistic classifier of low computational complexity, which can deal with multiclass problems. This research area deals with methods that solve problems in which labelled training data sets are assumed to be available only for a source domain, while classification is needed in the target domain with different, yet related characteristics. Classification takes place with a model of weight coefficients for hyperplanes which separate features in the transformed feature space. In term of logistic regression, our domain adaptation method adjusts the model parameters by iterative labelling of the target test data set. These labelled data features are iteratively added...
ISPRS Journal of Photogrammetry and Remote Sensing, 2014
ABSTRACT Many public and private decisions rely on geospatial information stored in a GIS databas... more ABSTRACT Many public and private decisions rely on geospatial information stored in a GIS database. For good decision making this information has to be complete, consistent, accurate and up-to-date. In this paper we introduce a new approach for the semi-automatic verification of a specific part of the, possibly outdated GIS database, namely cropland and grassland objects, using mono-temporal very high resolution (VHR) multispectral satellite images. The approach consists of two steps: first, a supervised pixel-based classification based on a Markov Random Field is employed to extract image regions which contain agricultural areas (without distinction between cropland and grassland), and these regions are intersected with boundaries of the agricultural objects from the GIS database. Subsequently, GIS objects labelled as cropland or grassland in the database and showing agricultural areas in the image are subdivided into different homogeneous regions by means of image segmentation, followed by a classification of these segments into either cropland or grassland using a Support Vector Machine. The classification result of all segments belonging to one GIS object are finally merged and compared with the GIS database label. The developed approach was tested on a number of images. The evaluation shows that errors in the GIS database can be significantly reduced while also speeding up the whole verification task when compared to a manual process.
ABSTRACT In this work we address the task of contextual classification of an airborne LiDAR point... more ABSTRACT In this work we address the task of contextual classification of an airborne LiDAR point cloud. For that purpose, we integrate a Random Forest classifier into a Conditional Random Field (CRF) framework. A CRF has been shown to deliver good results discerning multiple classes. It is a flexible approach for obtaining a reliable classification even in complex urban scenes. The incorporation of multi-scale features improves the results further. Based on the classification results, 2D building and tree objects are generated and evaluated by the benchmark of ISPRS WG III/4.
2012 IEEE International Geoscience and Remote Sensing Symposium, 2012
This paper gives an overview about advanced techniques for classification and object detection th... more This paper gives an overview about advanced techniques for classification and object detection that are being adopted for urban object detection from LiDAR data. The paper covers local supervised classifiers such as AdaBoost, SVM and Random Forests, statistical models of context such as Markov Random Fields and Conditional Random Fields, and sampling techniques. The relevance of features is also discussed. Applications include DTM generation and the extraction of buildings, trees, and low vegetation.
ISPRS Ann. Photogramm. Remote Sens. Spatial Inf. Sci., IV-2/W6, 2019
This paper presents a method for the classification of images of silk fabrics with the aim to pre... more This paper presents a method for the classification of images of silk fabrics with the aim to predict properties such as the place and time of origin and the production technique. The proposed method was developed in the context of the EU project SILKNOW (http://silknow.eu/). In the context of classification, we address the problem of limited as well as not fully labelled data and investigate the connection between the distinct variables. A pre-trained Convolutional Neural Network (CNN) is used for the feature extraction and a classification network realizing Multi-task learning (MTL) is trained based on these features. The training procedure is adapted to enable the consideration of images that do not have a label for all tasks. Additionally, MTL with fully labeled training data is investigated for the classification of silk fabrics. The impact of both MTL approaches is compared to single-task learning based on two different class structures. We achieve overall accuracies of 92-95% and average F1-scores of 88-90% in our best experiments.
ISPRS Annals of Photogrammetry, Remote Sensing and Spatial Information Sciences
Supervised machine learning needs high quality, densely sampled and labelled training data. Trans... more Supervised machine learning needs high quality, densely sampled and labelled training data. Transfer learning (TL) techniques have been devised to reduce this dependency by adapting classifiers trained on different, but related, (source) training data to new (target) data sets. A problem in TL is how to quantify the relatedness of a source quickly and robustly, because transferring knowledge from unrelated data can degrade the performance of a classifier. In this paper, we propose a method that can select a nearly optimal source from a large number of candidate sources. This operation depends only on the marginal probability distributions of the data, thus allowing the use of the often abundant unlabelled data. We extend this method to multi-source selection by optimizing a weighted combination of sources. The source weights are computed using a very fast boosting-like optimization scheme. The run-time complexity of our method scales linearly in regard to the number of candidate sou...
ISPRS - International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences
We propose a new approach for the automatic detection of network structures in raster data. The m... more We propose a new approach for the automatic detection of network structures in raster data. The model for the network structure is represented by a graph whose nodes and edges correspond to junction-points and to connecting line segments, respectively; nodes and edges are further described by certain parameters. We embed this model in the probabilistic framework of marked point processes and determine the most probable configuration of objects by stochastic sampling. That is, different graph configurations are constructed randomly by modifying the graph entity parameters, by adding and removing nodes and edges to/ from the current graph configuration. Each configuration is then evaluated based on the probabilities of the changes and an energy function describing the conformity with a predefined model. By using the Reversible Jump Markov Chain Monte Carlo sampler, a global optimum of the energy function is determined. We apply our method to the detection of river and tidal channel ne...
ISPRS Annals of Photogrammetry, Remote Sensing and Spatial Information Sciences, 2015
The fully automated analysis of 3D point clouds is of great importance in photogrammetry, remote ... more The fully automated analysis of 3D point clouds is of great importance in photogrammetry, remote sensing and computer vision. For reliably extracting objects such as buildings, road inventory or vegetation, many approaches rely on the results of a point cloud classification, where each 3D point is assigned a respective semantic class label. Such an assignment, in turn, typically involves statistical methods for feature extraction and machine learning. Whereas the different components in the processing workflow have extensively, but separately been investigated in recent years, the respective connection by sharing the results of crucial tasks across all components has not yet been addressed. This connection not only encapsulates the interrelated issues of neighborhood selection and feature extraction, but also the issue of how to involve spatial context in the classification step. In this paper, we present a novel and generic approach for 3D scene analysis which relies on…
ISPRS Annals of Photogrammetry, Remote Sensing and Spatial Information Sciences, 2015
Tracking-by-detection is a widely used practice in recent tracking systems. These usually rely on... more Tracking-by-detection is a widely used practice in recent tracking systems. These usually rely on independent single frame detections that are handled as observations in a recursive estimation framework. If these observations are imprecise the generated trajectory is prone to be updated towards a wrong position. In contrary to existing methods our novel approach uses a Dynamic Bayes Network in which the state vector of a recursive Bayes filter, as well as the location of the tracked object in the image are modelled as unknowns. These unknowns are estimated in a probabilistic framework taking into account a dynamic model, and a state-of-the-art pedestrian detector and classifier. The classifier is based on the Random Forest-algorithm and is capable of being trained incrementally so that new training samples can be incorporated at runtime. This allows the classifier to adapt to the changing appearance of a target and to unlearn outdated features. The approach is evaluated on a publicl...
ISPRS - International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, 2015
In this paper we address the problem of classification of remote sensing images in the framework ... more In this paper we address the problem of classification of remote sensing images in the framework of transfer learning with a focus on domain adaptation. The main novel contribution is a method for transductive transfer learning in remote sensing on the basis of logistic regression. Logistic regression is a discriminative probabilistic classifier of low computational complexity, which can deal with multiclass problems. This research area deals with methods that solve problems in which labelled training data sets are assumed to be available only for a source domain, while classification is needed in the target domain with different, yet related characteristics. Classification takes place with a model of weight coefficients for hyperplanes which separate features in the transformed feature space. In term of logistic regression, our domain adaptation method adjusts the model parameters by iterative labelling of the target test data set. These labelled data features are iteratively added...
ISPRS Journal of Photogrammetry and Remote Sensing, 2014
ABSTRACT Many public and private decisions rely on geospatial information stored in a GIS databas... more ABSTRACT Many public and private decisions rely on geospatial information stored in a GIS database. For good decision making this information has to be complete, consistent, accurate and up-to-date. In this paper we introduce a new approach for the semi-automatic verification of a specific part of the, possibly outdated GIS database, namely cropland and grassland objects, using mono-temporal very high resolution (VHR) multispectral satellite images. The approach consists of two steps: first, a supervised pixel-based classification based on a Markov Random Field is employed to extract image regions which contain agricultural areas (without distinction between cropland and grassland), and these regions are intersected with boundaries of the agricultural objects from the GIS database. Subsequently, GIS objects labelled as cropland or grassland in the database and showing agricultural areas in the image are subdivided into different homogeneous regions by means of image segmentation, followed by a classification of these segments into either cropland or grassland using a Support Vector Machine. The classification result of all segments belonging to one GIS object are finally merged and compared with the GIS database label. The developed approach was tested on a number of images. The evaluation shows that errors in the GIS database can be significantly reduced while also speeding up the whole verification task when compared to a manual process.
ABSTRACT In this work we address the task of contextual classification of an airborne LiDAR point... more ABSTRACT In this work we address the task of contextual classification of an airborne LiDAR point cloud. For that purpose, we integrate a Random Forest classifier into a Conditional Random Field (CRF) framework. A CRF has been shown to deliver good results discerning multiple classes. It is a flexible approach for obtaining a reliable classification even in complex urban scenes. The incorporation of multi-scale features improves the results further. Based on the classification results, 2D building and tree objects are generated and evaluated by the benchmark of ISPRS WG III/4.
2012 IEEE International Geoscience and Remote Sensing Symposium, 2012
This paper gives an overview about advanced techniques for classification and object detection th... more This paper gives an overview about advanced techniques for classification and object detection that are being adopted for urban object detection from LiDAR data. The paper covers local supervised classifiers such as AdaBoost, SVM and Random Forests, statistical models of context such as Markov Random Fields and Conditional Random Fields, and sampling techniques. The relevance of features is also discussed. Applications include DTM generation and the extraction of buildings, trees, and low vegetation.
Uploads