Descripteur
Termes IGN > informatique > intelligence artificielle > apprentissage automatique > apprentissage profond
apprentissage profond |
Documents disponibles dans cette catégorie (759)
![](./images/expand_all.gif)
![](./images/collapse_all.gif)
Etendre la recherche sur niveau(x) vers le bas
Use of artificial neural networks for selective omission in updating road networks / Qi Zhou in Cartographic journal (the), vol 51 n° 1 (February 2014)
![]()
[article]
Titre : Use of artificial neural networks for selective omission in updating road networks Type de document : Article/Communication Auteurs : Qi Zhou, Auteur ; Zhilin Li, Auteur Année de publication : 2014 Article en page(s) : pp 38 - 51 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Bases de données localisées
[Termes IGN] analyse comparative
[Termes IGN] carte de Kohonen
[Termes IGN] généralisation cartographique automatisée
[Termes IGN] mise à jour automatique
[Termes IGN] mise à jour cartographique
[Termes IGN] réseau neuronal artificiel
[Termes IGN] réseau routierRésumé : (Auteur) An important problem faced by national mapping agencies is frequent map updates. An ideal solution is only updating the large-scale map with other smaller scale maps undergoing automatic updates. This process may involve a series of operators, among which selective omission has received much attention. This study focuses on selective omission in a road network, and the use of an artificial neural network (i.e. a back propagation neural network, BPNN). The use of another type of artificial neural network (i.e. a self-organizing map, SOM) is investigated as a comparison. The use of both neural networks for selective omission is tested on a real-life road network. The use of a BPNN for practical application road updating is also tested. The results of selective omission are evaluated by overall accuracy. It is found that (1) the use of a BPNN can adaptively determine which and how many roads are to be retained at a specific scale, with an overall accuracy above 80%; (2) it may be hard to determine which and how many roads should be retained at a specific scale using an SOM. Therefore, the BPNN is more effective for selective omission in road updating. Numéro de notice : A2014-128 Affiliation des auteurs : non IGN Thématique : GEOMATIQUE/INFORMATIQUE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1179/1743277413Y.0000000042 En ligne : https://doi.org/10.1179/1743277413Y.0000000042 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=33033
in Cartographic journal (the) > vol 51 n° 1 (February 2014) . - pp 38 - 51[article]Exemplaires(1)
Code-barres Cote Support Localisation Section Disponibilité 030-2014011 RAB Revue Centre de documentation En réserve L003 Disponible Assessing the performance of two unsupervised dimensionality reduction techniques on hyperspectral APEX data for high resolution urban land-cover mapping / Luca Demarchi in ISPRS Journal of photogrammetry and remote sensing, vol 87 (January 2014)
![]()
[article]
Titre : Assessing the performance of two unsupervised dimensionality reduction techniques on hyperspectral APEX data for high resolution urban land-cover mapping Type de document : Article/Communication Auteurs : Luca Demarchi, Auteur ; Frank Canters, Auteur ; Claude Cariou, Auteur ; Giorgio Licciardi, Auteur ; Jonathan Cheung-Wai Chan, Auteur Année de publication : 2014 Article en page(s) : pp 166 - 179 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image optique
[Termes IGN] Airborne Prism Experiment
[Termes IGN] carte d'occupation du sol
[Termes IGN] classification par forêts d'arbres décisionnels
[Termes IGN] classification par réseau neuronal
[Termes IGN] classification par séparateurs à vaste marge
[Termes IGN] image APEX
[Termes IGN] image hyperspectrale
[Termes IGN] Perceptron multicoucheRésumé : (Auteur) Despite the high richness of information content provided by airborne hyperspectral data, detailed urban land-cover mapping is still a challenging task. An important topic in hyperspectral remote sensing is the issue of high dimensionality, which is commonly addressed by dimensionality reduction techniques. While many studies focus on methodological developments in data reduction, less attention is paid to the assessment of the proposed methods in detailed urban hyperspectral land-cover mapping, using state-of-the-art image classification approaches. In this study we evaluate the potential of two unsupervised data reduction techniques, the Autoassociative Neural Network (AANN) and the BandClust method – the first a transformation based approach, the second a feature-selection based approach – for mapping of urban land cover at a high level of thematic detail, using an APEX 288-band hyperspectral dataset. Both methods were tested in combination with four state-of-the-art machine learning classifiers: Random Forest (RF), AdaBoost (ADB), the multiple layer perceptron (MLP), and support vector machines (SVM). When used in combination with a strong learner (MLP, SVM) BandClust produces classification accuracies similar to or higher than obtained with the full dataset, demonstrating the method’s capability of preserving critical spectral information, required for the classifier to successfully distinguish between the 22 urban land-cover classes defined in this study. In the AANN data reduction process, on the other hand, important spectral information seems to be compromised or lost, resulting in lower accuracies for three of the four classifiers tested. Detailed analysis of accuracies at class level confirms the superiority of the SVM/Bandclust combination for accurate urban land-cover mapping using a reduced hyperspectral dataset. This study also demonstrates the potential of the new APEX sensor data for detailed mapping of land cover in spatially and spectrally complex urban areas. Numéro de notice : A2014-018 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1016/j.isprsjprs.2013.10.012 En ligne : https://doi.org/10.1016/j.isprsjprs.2013.10.012 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=32923
in ISPRS Journal of photogrammetry and remote sensing > vol 87 (January 2014) . - pp 166 - 179[article]Exemplaires(1)
Code-barres Cote Support Localisation Section Disponibilité 081-2014011 RAB Revue Centre de documentation En réserve L003 Disponible
Titre : Automatische Interpretation von Semantik aus digitalen Karten im World Wide Web Type de document : Thèse/HDR Auteurs : Fen Luo, Auteur ; Dieter Fritsch, Directeur de thèse ; Ralf Bill, Directeur de thèse Editeur : Stuttgart : Institüt für Photogrammetrie der Universität Stuttgart Année de publication : 2014 Note générale : bibliographie
Von der Fakultät Luft- und Raumfahrttechnik und Geodäsie der Universität Stuttgart zur Erlangung der Würde eines Doktors der Ingenieurwissenschaften (Dr.-Ing.) genehmigte AbhandlungLangues : Allemand (ger) Descripteur : [Vedettes matières IGN] Géomatique web
[Termes IGN] carte de Kohonen
[Termes IGN] données maillées
[Termes IGN] données vectorielles
[Termes IGN] échelle cartographique
[Termes IGN] exploration de données géographiques
[Termes IGN] objet géographique
[Termes IGN] réseau neuronal artificiel
[Termes IGN] toile d'araignée mondialeRésumé : (auteur) On the Internet there are innumerable spatial data representing different sections of the world in form of raster and vector maps. The information contained in these maps is not automatically discoverable, since it is encoded by means of certain map elements. Its semantics is not explicit unless interpreted by an observer. However, the map information can be interpreted by not only humans but also ma-chines. This already requires the large amount of data to be interpreted. We are going to summarize the automatic derivation of semantics from the maps in terms of automatic map interpretation. It involves a process of making the implicit information of a map inventory explicit. For this purpose we present the map interpretation as solutions.
The map interpretation of the current study is done with vector maps what can be found on the internet. For the targeted search of vector maps of the internet, a web crawler is specially developed. The web crawler is a search engine that specifically looks for vector maps. For this, exclusively the shapefile format is sought, which has become a standard format in the GIS environment and in which the vector maps are usually stored. In order to find shapefiles as many as possible, the search is carried out on servers where the probability of finding shapefiles is high. These servers were previously found through the keyword “shapefile download” by Google search.
The maps interpretation includes methods of interpretation of the map objects, of the map types, and of the map scale. First, we will introduce the method of interpreting the map objects. Our aim is to automatically detect the objects based on their specific characteristics. The object recognition is based on self-organizing map (SOM) that is borrowed from artificial intelligence. The map objects are clas-sified into, for example, building floor plan and road network. Its own characteristics should be found for each class and brought in one of the accessible forms of SOM, in this case, a parameter vector. The parameter vectors form the input patterns that are learned in the training phase of SOM. After the input patterns of all object classes of SOM have been learned, the parameter vector is evaluated for each of the present objects on the map and given to the SOM. By the previously successful learning of the input pattern, the objects can be assigned based on each of their calculated parameter vectors of the corresponding object class.
The interpretation of map type is presented as another method. Maps are categorized into different types according to their substantive content and purpose, such as river maps, road maps, contour maps, etc. As for the interpretation of objects, SOM is used here. Hence the input patterns will also be learned which represent the geometric characteristics of the map types. The characteristics arise from both the structure of individual objects and the topology between objects on a map. Now, with a given map in the SOM, the SOM recognizes the appropriate map type according to the learned input pattern. In addition, one obtains the filenames of the maps as well as the content of the website where the map was found. In the present thesis we also investigated how this additional information can help in the interpretation of map type.
The automatic interpretation of the map scale is a further method in addition to the interpretation of the map objects and map types, which is discussed in the present thesis. The interpretation of the map scale is implemented in two ways: the multi-representation and the details grade. In the former case, the scale of the relevant representation can be derived, where an identical object in different realistic representations on the map is shown; while in the latter case, the scale is derived from the details grade, on the basis of the fact that maps with different scales are displayed on different levels of details.Numéro de notice : 17347 Affiliation des auteurs : non IGN Thématique : GEOMATIQUE Nature : Thèse étrangère DOI : sans En ligne : http://doi.org/10.18419/opus-3960 Format de la ressource électronique : URL Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=83706 Comparison of support vector machine, neural network, and CART algorithms for the land-cover classification using limited training data points / Y. Shao in ISPRS Journal of photogrammetry and remote sensing, vol 70 (June 2012)
![]()
[article]
Titre : Comparison of support vector machine, neural network, and CART algorithms for the land-cover classification using limited training data points Type de document : Article/Communication Auteurs : Y. Shao, Auteur ; R. Lunetta, Auteur Année de publication : 2012 Article en page(s) : pp 78 - 87 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image optique
[Termes IGN] analyse comparative
[Termes IGN] analyse diachronique
[Termes IGN] classification dirigée
[Termes IGN] classification et arbre de régression
[Termes IGN] classification par réseau neuronal
[Termes IGN] classification par séparateurs à vaste marge
[Termes IGN] image Terra-MODIS
[Termes IGN] occupation du sol
[Termes IGN] Perceptron multicouche
[Termes IGN] série temporelleRésumé : (Auteur) Support vector machine (SVM) was applied for land-cover characterization using MODIS time-series data. Classification performance was examined with respect to training sample size, sample variability, and landscape homogeneity (purity). The results were compared to two conventional nonparametric image classification algorithms: multilayer perceptron neural networks (NN) and classification and regression trees (CART). For 2001 MODIS time-series data, SVM generated overall accuracies ranging from 77% to 80% for training sample sizes from 20 to 800 pixels per class, compared to 67–76% and 62–73% for NN and CART, respectively. These results indicated that SVM’s had superior generalization capability, particularly with respect to small training sample sizes. There was also less variability of SVM performance when classification trials were repeated using different training sets. Additionally, classification accuracies were directly related to sample homogeneity/heterogeneity. The overall accuracies for the SVM algorithm were 91% (Kappa = 0.77) and 64% (Kappa = 0.34) for homogeneous and heterogeneous pixels, respectively. The inclusion of heterogeneous pixels in the training sample did not increase overall accuracies. Also, the SVM performance was examined for the classification of multiple year MODIS time-series data at annual intervals. Finally, using only the SVM output values, a method was developed to directly classify pixel purity. Approximately 65% of pixels within the Albemarle–Pamlico Basin study area were labeled as “functionally homogeneous” with an overall classification accuracy of 91% (Kappa = 0.79). The results indicated a high potential for regional scale operational land-cover characterization applications. Numéro de notice : A2012-290 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1016/j.isprsjprs.2012.04.001 En ligne : https://doi.org/10.1016/j.isprsjprs.2012.04.001 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=31736
in ISPRS Journal of photogrammetry and remote sensing > vol 70 (June 2012) . - pp 78 - 87[article]Exemplaires(1)
Code-barres Cote Support Localisation Section Disponibilité 081-2012041 SL Revue Centre de documentation Revues en salle Disponible
Titre : 3D visualization of multivariate data Type de document : Thèse/HDR Auteurs : Harald Sanftmann, Auteur ; Daniel Weiskopf, Directeur de thèse ; H. Hauser, Directeur de thèse Editeur : Stuttgart : University of Stuttgart Année de publication : 2012 Importance : 157 p. Format : 21 x 30 cm Note générale : Bibliographie
Von der Fakultät Informatik, Elektrotechnik und Informationstechnik der Universität Stuttgart zur Erlangung der Würde eines Doktors der Naturwissenschaften, genehmigte AbhandlungLangues : Anglais (eng) Descripteur : [Termes IGN] anaglyphe
[Termes IGN] analyse multivariée
[Termes IGN] apprentissage profond
[Termes IGN] arbre de décision
[Termes IGN] éclairement lumineux
[Termes IGN] restitution numérique
[Termes IGN] semis de points
[Termes IGN] traitement de semis de points
[Termes IGN] visualisation 3D
[Vedettes matières IGN] GéovisualisationIndex. décimale : THESE Thèses et HDR Résumé : (auteur) Nowadays large amounts of data are organized in tables, especially in relational databases where the rows store the data items to which multiple attributes are stored in the columns. Information stored this way, having multiple (more than two or three) attributes, can be treated as multivariate data. Therefore, visualization methods for multivariate data have a large application area and high potential utility. This thesis focuses on the application of 3D scatter plots for the visualization of multivariate data. When dealing with 3D, spatial perception needs to be exploited, by effectively using depth cues to convey spatial information to the user. To improve the presentation of individual 3D scatter plots, a technique is presented that applies illumination to them, thus using the shape-from-shading depth cue. To enable the analysis not only of 3D but of multivariate data, a novel technique is introduced that allows the navigation between 3D scatter plots. Inspecting the large number of 3D scatter plots that can be projected from a multivariate data set is very time consuming. The analysis of multivariate data can benefit from automatic machine learning approaches. A presented method uses decision trees to increase the speed a user can gain an understanding of the multivariate data at no extra cost. Stereopsis can also support the display of 3D scatter plots. Here an improved anaglyph rendering technique is presented, significantly reducing ghosting artifacts. The technique is not only applicable for information visualization, but for general rendering or to present stereoscopic image data. Some information visualization algorithms require high computation time. Many of these algorithms can be parallelized to run interactively. A framework that supports the parallelization on shared and distributed memory systems is presented. Note de contenu : Introduction
1 - The Notion of 3D in Information Visualization
2 - Improving Depth Perception of 3D Scatter Plots
3 - 3D Scatter Plot Navigation
4 - Visualization with Decision Trees
5 - Anaglyph Stereo without Ghosting
6 - Distributed Visualization
Conclusion and OutlookNuméro de notice : 21571 Affiliation des auteurs : non IGN Thématique : GEOMATIQUE/IMAGERIE Nature : Thèse étrangère Note de thèse : PhD dissertation : Informatik, Elektrotechnik und Informationstechnik : Universität Stuttgart : 2012 DOI : 10.18419/opus-6401 En ligne : http://doi.org/10.18419/opus-6401 Format de la ressource électronique : URL Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=90561 An assessment of internal neural network parameters affecting image classification accuracy / L. Zhou in Photogrammetric Engineering & Remote Sensing, PERS, vol 77 n° 12 (December 2011)
PermalinkObject-based image analysis of high-resolution satellite images using modified cloud basis function neural network and probabilistic relaxation labeling process / A. Rizvi in IEEE Transactions on geoscience and remote sensing, vol 49 n° 12 Tome 1 (December 2011)
PermalinkDevelopment of a modified neural network-based land cover classification system using automated data selector and multiresolution remotely sensed data / S. Khorram in Geocarto international, vol 26 n° 6 (October 2011)
PermalinkSimilarity weighted instance-based learning for the generation of transition potentials in land use change modeling / F. Sangermano in Transactions in GIS, vol 14 n° 5 (October 2010)
PermalinkA neural network-based method for solving "nested hierarchy" areal interpolation problems / D. Merwin in Cartography and Geographic Information Science, vol 36 n° 4 (October 2009)
PermalinkPotentiality of feed-forward neural networks for classifying dark formations to oil spills and look-alikes / Konstantinos Topouzelis in Geocarto international, vol 24 n° 3 (June - July 2009)
PermalinkRepresenting geographical objects with scale-induced indeterminate boundaries: a neural network-based data model / José L. Silvan-Cardenas in International journal of geographical information science IJGIS, vol 23 n°3-4 (march - april 2009)
Permalinkvol 74 n° 10 - October 2008 - Artificial intelligence in remote sensing (Bulletin de Photogrammetric Engineering & Remote Sensing, PERS) / American society for photogrammetry and remote sensing
PermalinkUsing neural networks and cellular automata for modelling intra-urban land-use dynamics / C.M. Almeida in International journal of geographical information science IJGIS, vol 22 n° 8-9 (august 2008)
PermalinkThe application of artificial neural networks to the analysis of remotely sensed data / J.F. Mas in International Journal of Remote Sensing IJRS, vol 29 n°3-4 (February 2008)
Permalink