Image denoising using contextual modeling of curvelet coefficients

R. Kechichian 1 Carole Amiot 2 Christian Girard Jérémie Pescatore 3 Jocelyn Chanussot 4, 5 Michel Desvignes 6
1 Images et Modèles
CREATIS - Centre de Recherche en Acquisition et Traitement de l'Image pour la Santé
4 GIPSA-SIGMAPHY - SIGMAPHY
GIPSA-DIS - Département Images et Signal
6 GIPSA-AGPIG - AGPIG
GIPSA-DIS - Département Images et Signal
Abstract : We propose an image denoising method which takes curvelet domain inter-scale, inter-location and inter-orientation dependencies into account in a maximum a posteriori labeling of the curvelet coefficients of a noisy image. The rationale is that generalized neighborhoods of curvelet coefficients contain more reliable information on the true image than individual coefficients. Based on the labeling of coefficients and their magnitudes, a smooth thresholding functional produces denoised coefficients from which the denoised image is reconstructed. We also outline a faster approach to labeling and thresholding, relying on contextual comparisons of coefficients. Quantitative and qualitative evaluations on natural and X-ray images show that our method outperforms related multiscale approaches and compares favorably to the state-of-art BM3D method on X-ray data while executing faster.
Type de document :
Communication dans un congrès
21st IEEE International Conference on Image Processing (ICIP 2014), Oct 2014, Paris, France. pp.2659-2663, 〈10.1109/ICIP.2014.7025538〉
Liste complète des métadonnées

http://hal.univ-grenoble-alpes.fr/hal-01128450
Contributeur : Vincent Couturier-Doux <>
Soumis le : lundi 9 mars 2015 - 17:14:49
Dernière modification le : mardi 10 juillet 2018 - 01:17:49

Identifiants

Citation

R. Kechichian, Carole Amiot, Christian Girard, Jérémie Pescatore, Jocelyn Chanussot, et al.. Image denoising using contextual modeling of curvelet coefficients. 21st IEEE International Conference on Image Processing (ICIP 2014), Oct 2014, Paris, France. pp.2659-2663, 〈10.1109/ICIP.2014.7025538〉. 〈hal-01128450〉

Partager

Métriques

Consultations de la notice

379