关键词: Contourlet变换, 全变差, 细节保持, 去噪

Abstract: Image denoising with details sufficiently preserved has important significance in image processing field.A novel hybrid denoising algorithm is proposed，which combines the contourlet shrinkage method with the total variation approach.Using the spatially adaptive total variation，the difference image between the noisy image and the contourlet hard-threshold shrinkage image is filtered.The filtered difference image is added back to the shrinkage image to get the final denoised image.Experimental results show that，compared with existing typical denoising methods，the proposed algorithm can effectively remove the noise and Gibbs-like artifacts，and has superior performance in preserving the important detail information，such as edges and textures.

Key words: Contourlet transform, total variation, detail-preserving, denoising