A Numerical Model for Large Deformation on Meteorological Images

Jean-Paul Berroir 1 Isabelle Herlin 1 Isaac Cohen 1
1 AIR - Image Processing and Dynamic and Satellite Data
Inria Paris-Rocquencourt
Abstract : The purpose of non-rigid motion in computer vision is the study of the geometrical deformation that may affect an object during its temporal evolution. One often assumes that this process is local and continuous; it is therefore the case for local geometrical features, which are considered as relevant for studying deformation. This assumption is no more valid when the structure is widely distorted between two successive temporal occurrences, and information about the underlying physical phenomenon is then required to analyze motion. In this paper, we propose a geometrical model of evolution that may be viewed as an approximation of the true physical deformation: we are reconstructing the surface generated by the evolution of the boundary of the structure. We present a stable numerical implementation of this model and apply it to the tracking of a vortex on a sequence of meteorological images. Fulltext Preview
Type de document :
Communication dans un congrès
Springer Berlin Heidelberg. International Conference on Analysis and Optimisation of Systems, Images, Wavelets and PDE's (ICAOS), Jun 1996, Paris, France. 219, pp.92-99, 1996, Lecture Notes in Control and Information Sciences. 〈10.1007/3-540-76076-8_121〉
Liste complète des métadonnées
https://hal.inria.fr/inria-00532686
Contributeur : Brigitte Briot <>
Soumis le : jeudi 4 novembre 2010 - 13:38:49
Dernière modification le : vendredi 25 mai 2018 - 12:02:05

Lien texte intégral

openaccess

Identifiants

Collections

INRIA | INRIA-ROCQ

Citation

Jean-Paul Berroir, Isabelle Herlin, Isaac Cohen. A Numerical Model for Large Deformation on Meteorological Images. Springer Berlin Heidelberg. International Conference on Analysis and Optimisation of Systems, Images, Wavelets and PDE's (ICAOS), Jun 1996, Paris, France. 219, pp.92-99, 1996, Lecture Notes in Control and Information Sciences. 〈10.1007/3-540-76076-8_121〉. 〈inria-00532686〉

Exporter

BibTeX TEI DC DCterms EndNote

Partager

Métriques

Consultations de la notice

45

Contact


archive-ouverte@inria.fr