Descripteur
Termes IGN > mathématiques > statistique mathématique > analyse de données > classification > classification par séparateurs à vaste marge
classification par séparateurs à vaste margeSynonyme(s)classification SVMVoir aussi |
Documents disponibles dans cette catégorie (254)
Ajouter le résultat dans votre panier
Visionner les documents numériques
Affiner la recherche Interroger des sources externes
Etendre la recherche sur niveau(x) vers le bas
Titre : Tree species classification with multiple source remote sensing data Type de document : Thèse/HDR Auteurs : Eetu Puttonen, Auteur ; Juha Hyyppä, Directeur de thèse Editeur : Helsinki : Finnish Geodetic Institute FGI Année de publication : 2012 Collection : Publications of the Finnish Geodetic Institute, ISSN 0085-6932 num. 145 Importance : 86 p. Format : 21 x 30 cm ISBN/ISSN/EAN : 978-951-711-289-5 Note générale : Doctoral dissertation University of Helsinki, Faculty of Science, Department of Physics, geophysics and astronomy Finnish Geodetic Institute, Department of Photogrammetry and Remote Sensing
ISBN du pdfLangues : Anglais (eng) Descripteur : [Vedettes matières IGN] Applications photogrammétriques
[Termes IGN] analyse discriminante
[Termes IGN] arbre (flore)
[Termes IGN] capteur hyperspectral
[Termes IGN] classification
[Termes IGN] classification par séparateurs à vaste marge
[Termes IGN] données lidar
[Termes IGN] données localisées 3D
[Termes IGN] espèce végétale
[Termes IGN] fusion de données
[Termes IGN] image spectrale
[Termes IGN] semis de pointsRésumé : (auteur) Remote sensing is a study that provides information on targets of interest without direct interaction with them. Generally, the term is used for measurement techniques that detect electro-magnetic radiation emitted or reflected from the targets.
Commonly used wavelength ranges include visible, infra-red, microwaves, and thermal bands. This information can be exploited to determine the structural and spectral properties of targets. Remote sensing techniques are typically utilized in mapping solutions, environment monitoring, target recognition, change detection, and in creation of physical models.
In Finland, remote sensing research is of specific importance in forest sciences and industry as they need precise information on tree quantity and quality over large forest ranges. Tree species information on individual tree level is an important parameter to achieve this goal.
The aim of this thesis is to study how individual tree species information can be extracted with multiple source remote sensing data. The aim is achieved by combining spatial and spectral remote sensing data. Structural properties of individual trees are determined from three dimensional point clouds collected with laser scanners. Spectral properties of trees are collected with cameras or spectrometers.
The thesis consists of four separate studies. The first study examined how shading information of trees canopies could be exploited to improve tree species classification in data collected with airborne sensors. The second study examined the classification performance of a low-cost, multi-sensor, mobile mapping system. The third study investigated the classification performance and accuracy of a novel, active hyperspectral laser scanner. Finally, the fourth study evaluated the suitability of artificial surfaces as on-site intensity calibration targets.
The results of the three classification studies showed that the use of combined point cloud and spectral information yielded the best classification results in all study cases when compared against classification results obtained with only structural or spectral information. Moreover, the studies showed that the improved results could be achieved with a low total number of mixed structural and spectral classification parameters. The fourth study showed that the artificial surfaces work as calibration surfaces only in limited cases.
The main outcome of the thesis was that the active remote sensing systems measuring multiple wavelengths simultaneously should be promoted. They have a significant potential to improve tree species classification performance even with a few application-specific wavelengths.Numéro de notice : 15863 Affiliation des auteurs : non IGN Thématique : FORET/IMAGERIE Nature : Thèse étrangère Note de thèse : Doctoral dissertation : Photogrammetry and Remote Sensing : University of Helsinki : 2012 En ligne : http://hdl.handle.net/10138/33956 Format de la ressource électronique : URL Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=93650 Relevance assessment of full-waveform lidar data for urban area classification / Clément Mallet in ISPRS Journal of photogrammetry and remote sensing, vol 66 n° 6 supplement (December 2011)
[article]
Titre : Relevance assessment of full-waveform lidar data for urban area classification Type de document : Article/Communication Auteurs : Clément Mallet , Auteur ; Frédéric Bretar, Auteur ; Michel Roux, Auteur ; Uwe Soergel, Auteur ; Christian Heipke, Auteur Année de publication : 2011 Article en page(s) : pp 71 - 84 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Lasergrammétrie
[Termes IGN] analyse comparative
[Termes IGN] classification automatique
[Termes IGN] classification par séparateurs à vaste marge
[Termes IGN] données lidar
[Termes IGN] lidar à retour d'onde complète
[Termes IGN] méthode des moindres carrés
[Termes IGN] pertinence
[Termes IGN] semis de points
[Termes IGN] zone urbaineRésumé : (Auteur) Full-waveform lidar data are increasingly being available. Morphological features can be retrieved from the echoes composing the waveforms, and are now extensively used for a large variety of land-cover mapping issues. However, the genuine contribution of these features with respect to those computed from standard discrete return lidar systems has been barely theoretically investigated. This paper therefore aims to study the potential of full-waveform data through the automatic classification of urban areas in building, ground, and vegetation points. Two waveform processing methods, namely a non-linear least squares method and a marked point process approach, are used to fit the echoes both with symmetric and asymmetric modeling functions. The performance of the extracted full-waveform features for the classification problem are then compared to a large variety of multiple-pulse features thanks to three feature selection methods. A support vector machines classifier is finally used to label the point cloud according to various scenarios based on the rank of the features. This allows to find the best classification strategy as well as the minimal feature subsets allowing to achieve the highest classification accuracy possible for each of the three feature selection methods. The results show that the echo amplitude as well as two features computed from the radiometric calibration of full-waveform data, namely the cross-section and the backscatter coefficient, significantly contribute to the high classification accuracies reported in this paper (around 95%). Conversely, features extracted from the non Gaussian modelling of the echoes are not relevant for the discrimination of vegetation, ground, and buildings in urban areas. Numéro de notice : A2011-520 Affiliation des auteurs : IGN+Ext (1940-2011) Thématique : IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1016/j.isprsjprs.2011.09.008 Date de publication en ligne : 12/10/2011 En ligne : http://www.sciencedirect.com/science/article/pii/S0924271611001055 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=31414
in ISPRS Journal of photogrammetry and remote sensing > vol 66 n° 6 supplement (December 2011) . - pp 71 - 84[article]Exemplaires(1)
Code-barres Cote Support Localisation Section Disponibilité 081-2011071 SL Revue Centre de documentation Revues en salle Disponible SVM-based unmixing-to-classification conversion for hyperspectral abundance quantification / F. Mianji in IEEE Transactions on geoscience and remote sensing, vol 49 n° 11 Tome 1 (November 2011)
[article]
Titre : SVM-based unmixing-to-classification conversion for hyperspectral abundance quantification Type de document : Article/Communication Auteurs : F. Mianji, Auteur ; Y. Zhang, Auteur Année de publication : 2011 Article en page(s) : pp 4318 - 4327 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image optique
[Termes IGN] analyse des mélanges spectraux
[Termes IGN] analyse infrapixellaire
[Termes IGN] classification par séparateurs à vaste marge
[Termes IGN] image hyperspectrale
[Termes IGN] signature spectraleRésumé : (Auteur) Need for a priori knowledge of the components comprising each pixel in a scene has set the endmember determination, rather than the endmember abundance quantification, as the primary focus of many unmixing approaches. In the absence of the information about the pure signatures present in an image scene, which is often the case, the mean spectra of the pixel vectors, directly extracted from the scene, are usually used as the pure signatures' spectra. This approach which is mathematically optimized for unmixing problems with a priori known information ignores some statistical properties of the extracted samples and leads to a suboptimal solution for real situations. This paper proposes a novel learning-based unmixing-to-classification conversion model to treat the abundance quantification task as a classification problem. Support vector machine, as an efficient classifier, is used to realize this model. It exploits the statistical nature (endmember spectral variability) of the extracted endmember representatives from the hyperspectral scene, rather than solving the problem according to the ideal model in which only the mean spectra of each training sample set is used. Several experiments are carried out on simulated and real hyperspectral images. The obtained results validate the high performance of the proposed technique in abundance quantification which is a key subpixel information detection capability. Numéro de notice : A2011-446 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern DOI : 10.1109/TGRS.2011.2166766 Date de publication en ligne : 06/10/2011 En ligne : https://doi.org/10.1109/TGRS.2011.2166766 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=31224
in IEEE Transactions on geoscience and remote sensing > vol 49 n° 11 Tome 1 (November 2011) . - pp 4318 - 4327[article]Exemplaires(1)
Code-barres Cote Support Localisation Section Disponibilité 065-2011111A RAB Revue Centre de documentation En réserve L003 Disponible Damage assessment of 2010 Haïti earthquake with post-earthquake satellite image by support vector selection and adaptation / Gülsen Taskin Kaya in Photogrammetric Engineering & Remote Sensing, PERS, vol 77 n° 10 (October 2011)
[article]
Titre : Damage assessment of 2010 Haïti earthquake with post-earthquake satellite image by support vector selection and adaptation Type de document : Article/Communication Auteurs : Gülsen Taskin Kaya, Auteur ; Nebiye Musaoglu, Auteur ; Okan Ersoy, Auteur Année de publication : 2011 Article en page(s) : pp 1025 - 1035 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Applications de télédétection
[Termes IGN] analyse comparative
[Termes IGN] carte thématique
[Termes IGN] catastrophe naturelle
[Termes IGN] classification par séparateurs à vaste marge
[Termes IGN] dommage matériel
[Termes IGN] Haïti
[Termes IGN] image multitemporelle
[Termes IGN] image satellite
[Termes IGN] séismeRésumé : (Auteur) Remote sensing technology is a powerful tool to extract regions damaged after an earthquake. There are two methodological approaches in detection of earthquake damage: mono-temporal and multi-temporal. Especially for providing effective emergency management, the mono-temporal approach is generally preferred in extraction of earthquake damage as it does not depend on availability of pre-earthquake imagery. For this purpose, a novel method called support vector selection and adaptation (svsa) has been introduced to detect the damaged regions from a post-earthquake image. In this study, the SVSA method was applied to the region where the Haiti Presidential Palace and Cathedral is located, and the damaged regions were identified. The performance of the SVSA method in identification of the damaged regions was evaluated by comparing the thematic maps obtained by classifying pre- and post-earthquake images. Additionally, the damage patterns for the city of Port-au-Prince were estimated by the SVSA. Numéro de notice : A2011-433 Affiliation des auteurs : non IGN Thématique : IMAGERIE Nature : Article DOI : 10.14358/PERS.77.10.1025 En ligne : https://doi.org/10.14358/PERS.77.10.1025 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=31211
in Photogrammetric Engineering & Remote Sensing, PERS > vol 77 n° 10 (October 2011) . - pp 1025 - 1035[article]Full waveform-based analysis for forest type information derivation from large footprint spaceborne lidar data / Junjie Zhang in Photogrammetric Engineering & Remote Sensing, PERS, vol 77 n° 3 (March 2011)
[article]
Titre : Full waveform-based analysis for forest type information derivation from large footprint spaceborne lidar data Type de document : Article/Communication Auteurs : Junjie Zhang, Auteur ; A. De Gier, Auteur ; Y. Xing, Auteur ; Gunho Sohn, Auteur Année de publication : 2011 Conférence : SilviLaser 2010, 10th International Conference on LiDAR Applications for Assessing Forest Ecosystems 14/09/2010 17/09/2010 Fribourg Allemagne Article en page(s) : pp 281 - 290 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Lasergrammétrie
[Termes IGN] canopée
[Termes IGN] capteur spatial
[Termes IGN] classification par séparateurs à vaste marge
[Termes IGN] couvert forestier
[Termes IGN] décomposition de Gauss
[Termes IGN] données lidar
[Termes IGN] feuillu
[Termes IGN] lidar à retour d'onde complète
[Termes IGN] Pinophyta
[Termes IGN] signal laserRésumé : (Auteur) This study developed a new method to derive forest type information from large-footprint lidar data based on full waveform analysis. For this purpose, the raw waveform was decomposed into Gaussian components, and canopy return and ground return of the waveforms were separated. Two types of metrics hypothesized to have relationship with forest types were derived from the canopy return part of the waveform. The first type of metrics is quantile-based metrics reflecting the vertical distribution of canopy return energy, and the second type is statistical characteristics of the Gaussian components of canopy return part. Support Vector Machine classification was applied to different combinations of the metrics to find their relationship with different forest types. The results showed that the second type of metrics, indicating the canopy stratum characteristics, showed great promise in separating broad-leaved and needle-leaved forests with the accuracy ranging from 88.68 percent to 90.57 percent and Kappa statistic from 0.7406 to 0.7868. Numéro de notice : A2011-081 Affiliation des auteurs : non IGN Thématique : FORET/IMAGERIE Nature : Article DOI : 10.14358/PERS.77.3.281 En ligne : https://doi.org/10.14358/PERS.77.3.281 Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=30862
in Photogrammetric Engineering & Remote Sensing, PERS > vol 77 n° 3 (March 2011) . - pp 281 - 290[article]Parameterizing support vector machines for land cover classification / X. Yang in Photogrammetric Engineering & Remote Sensing, PERS, vol 77 n° 1 (January 2011)PermalinkUncertainty analysis for the classification of multispectral satellite images using SVMs and SOMs / F. Giacco in IEEE Transactions on geoscience and remote sensing, vol 48 n° 10 (October 2010)PermalinkSemisupervised one-class support vector machine for classification of remote sensing data / Jordi Munoz-Mari in IEEE Transactions on geoscience and remote sensing, vol 48 n° 8 (August 2010)PermalinkLand-cover change detection using one-class support vector machine / P. Li in Photogrammetric Engineering & Remote Sensing, PERS, vol 76 n° 3 (March 2010)PermalinkAnalyse de données lidar à retour d'onde complète pour la classification en milieu urbain = Analysis of Full-Waveform lidar data for urban area mapping / Clément Mallet (2010)PermalinkSupport vector machines for urban growth modeling / B. Huang in Geoinformatica, vol 14 n° 1 (January 2010)PermalinkPermalinkTerrain surfaces and 3-D Landcover classification from small footprint full-waveform Lidar data: application to badlands / Frédéric Bretar in Hydrology and Earth System Sciences, HESS, vol 13 n° 8 (26/08/2009)PermalinkApprentissage automatique des classes d'occupation du sol et représentation en mots visuels des images satellitaires / Marie Lauginie Lienou (2009)PermalinkDétection et caractérisation de la végétation en milieu urbain à partir d'images aériennes haute résolution / Corina Iovan (2009)Permalink