Parameter estimation for marked point processes. Application to object extraction from remote sensing images

Florent Chatelain 1 Xavier Descombes 2 Josiane Zerubia 2
1 GIPSA-SA-IGA - SA-IGA
GIPSA-DA - Département Automatique, GIPSA-DIS - Département Images et Signal
2 ARIANA - Inverse problems in earth monitoring
CRISAM - Inria Sophia Antipolis - Méditerranée , SIS - Signal, Images et Systèmes
Abstract : This communication addresses the problem of estimating the parameters of a family of marked point processes. These processes are of interest in extraction of object networks from remote sensing images. They are defined from a combination of several energy terms. First, a data energy term controls the localization of the objects with respect to the data. Second, prior information is given by intern energy terms corresponding to geometrical constraints on the configuration to be detected. An estimation procedure of the weight associated with these energies is studied. The application to unsupervised detection of objects is finally discussed.
Type de document :
Communication dans un congrès
D. Cremers et al. 7th International Conference on Energy Minimization Methods in Computer Vision and Pattern Recognition., Aug 2009, Bonn, Germany. Springer-Verlag Berlin Heidelberg, 5681, pp.221-234, 2009, Lecture Notes in Computer Science (LNCS)
Liste complète des métadonnées

https://hal.archives-ouvertes.fr/hal-00391719
Contributeur : Florent Chatelain <>
Soumis le : jeudi 4 juin 2009 - 15:26:14
Dernière modification le : samedi 23 janvier 2016 - 01:13:07

Identifiants

  • HAL Id : hal-00391719, version 1

Collections

Citation

Florent Chatelain, Xavier Descombes, Josiane Zerubia. Parameter estimation for marked point processes. Application to object extraction from remote sensing images. D. Cremers et al. 7th International Conference on Energy Minimization Methods in Computer Vision and Pattern Recognition., Aug 2009, Bonn, Germany. Springer-Verlag Berlin Heidelberg, 5681, pp.221-234, 2009, Lecture Notes in Computer Science (LNCS). <hal-00391719>

Partager

Métriques

Consultations de la notice

247