Bayesian non-local means filter, image redundancy and adaptive dictionaries for noise removal

Charles Kervrann 1, 2 Jérôme Boulanger 3, 1 Pierrick Coupé 4
2 VISTA - Vision spatio-temporelle et active
IRISA - Institut de Recherche en Informatique et Systèmes Aléatoires, Inria Rennes – Bretagne Atlantique
3 BUNRAKU - Perception, decision and action of real and virtual humans in virtual environments and impact on real environments
IRISA - Institut de Recherche en Informatique et Systèmes Aléatoires, ENS Cachan - École normale supérieure - Cachan, Inria Rennes – Bretagne Atlantique
4 VisAGeS - Vision, Action et Gestion d'informations en Santé
INSERM - Institut National de la Santé et de la Recherche Médicale : U746, Inria Rennes – Bretagne Atlantique , IRISA-D5 - SIGNAUX ET IMAGES NUMÉRIQUES, ROBOTIQUE
Abstract : Partial Differential equations (PDE), wavelets-based methods and neighborhood filters were proposed as locally adaptive machines for noise removal. Recently, Buades, Coll and Morel proposed the Non-Local (NL-) means filter for image denoising. This method replaces a noisy pixel by the weighted average of other image pixels with weights reflecting the similarity between local neighborhoods of the pixel being processed and the other pixels. The NL-means filter was proposed as an intuitive neighborhood filter but theoretical connections to diffusion and non-parametric estimation approaches are also given by the authors. In this paper we propose another bridge, and show that the NL-means filter also emerges from the Bayesian approach with new arguments. Based on this observation, we show how the performance of this filter can be significantly improved by introducing adaptive local dictionaries and a new statistical distance measure to compare patches. The new Bayesian NL-means filter is better parametrized and the amount of smoothing is directly determined by the noise variance (estimated from image data) given the patch size. Experimental results are given for real images with artificial Gaussian noise added, and for images with real image-dependent noise.
Type de document :
Communication dans un congrès
SSVM'07 Proceedings of the 1st international conference on Scale space and variational methods in computer vision, May 2007, Ischia, Italy. 4485, pp.520-532, 2007
Liste complète des métadonnées

Littérature citée [39 références]  Voir  Masquer  Télécharger

https://hal.archives-ouvertes.fr/hal-00645444
Contributeur : Pierrick Coupé <>
Soumis le : lundi 28 novembre 2011 - 10:14:37
Dernière modification le : mercredi 16 mai 2018 - 11:23:16
Document(s) archivé(s) le : vendredi 16 novembre 2012 - 12:10:55

Fichier

2007_ssvm_kervrann.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : hal-00645444, version 1

Citation

Charles Kervrann, Jérôme Boulanger, Pierrick Coupé. Bayesian non-local means filter, image redundancy and adaptive dictionaries for noise removal. SSVM'07 Proceedings of the 1st international conference on Scale space and variational methods in computer vision, May 2007, Ischia, Italy. 4485, pp.520-532, 2007. 〈hal-00645444〉

Partager

Métriques

Consultations de la notice

506

Téléchargements de fichiers

851