Skip to Main content Skip to Navigation
Theses

Une technique de relaxation pour la mise en correspondance d'images: Application à la reconnaissance d'objets et au suivi du visage.

Abstract : Local invariant features are a powerful tool for finding correspondences between images since they are robust to cluttered background, occlusion and viewpoint changes. However, they suffer the lack of global information and fail to resolve ambiguities that can occur when an image has multiple similar regions. In the first part of this thesis, we describe a matching algorithm based on a relaxation scheme, which makes use of contextual information for better performances. We show how the relaxation scheme can be made robust and fast, and we apply it in the case of object recognition. In the second part of this thesis, we tackle the problem of face detection and tracking in video sequences. We propose a simple and eficient face detection method in color images, and show how the matching method described in the first part can be used for tracking faces in video sequences.
Document type :
Theses
Complete list of metadatas

Cited literature [171 references]  Display  Hide  Download

https://tel.archives-ouvertes.fr/tel-00263567
Contributor : Cécile Chabanne <>
Submitted on : Wednesday, March 12, 2008 - 2:41:16 PM
Last modification on : Wednesday, June 24, 2020 - 4:18:07 PM
Long-term archiving on: : Friday, September 28, 2012 - 11:05:24 AM

Identifiers

  • HAL Id : tel-00263567, version 1

Collections

Citation

Dro Désiré Sidibe. Une technique de relaxation pour la mise en correspondance d'images: Application à la reconnaissance d'objets et au suivi du visage.. Interface homme-machine [cs.HC]. Université Montpellier II - Sciences et Techniques du Languedoc, 2007. Français. ⟨tel-00263567⟩

Share

Metrics

Record views

429

Files downloads

1348