Modification of the edge stopping function in anistropic diffusion process - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès RITS 2011:Congrès recherche en imagerie et technologie pour la santé Année : 2011

Modification of the edge stopping function in anistropic diffusion process

Résumé

In this paper we use an anisotropic diffusion process formulated in a level set framework using robust statistics. We show that smoothing, while preserving edges, helps the image segmentation of glenohumeral joint of shoulder for MRI data. The equation is based on an automatic edge stopping function that controls image smoothing. The link between robust statistics and anisotropic diffusion is used from the relationships between the edge stopping function, the influence function and the robust error norm of the estimator. We perform a comparison between three edge stopping functions, and then we propose to replace the Leclerc function already used in our diffusion equation by the Tukey's biweight function. Then we use the entropy criterion to evaluate our results and approve our choice
Fichier non déposé

Dates et versions

hal-00725546 , version 1 (27-08-2012)

Identifiants

  • HAL Id : hal-00725546 , version 1

Citer

Refka Ragoubi Hor, Valérie Burdin, Taha Jerbi, Olivier Rémy-Néris. Modification of the edge stopping function in anistropic diffusion process. RITS 2011:Congrès recherche en imagerie et technologie pour la santé, Apr 2011, Rennes, France. ⟨hal-00725546⟩
87 Consultations
0 Téléchargements

Partager

Gmail Facebook X LinkedIn More