Group sparsity residual constraint for image denoising with external nonlocal self-similarity prior

NEUROCOMPUTING

卷:275

Article

Zha, Zhiyuan¹  Zhang, Xinggan¹  Wang, Qiong¹  Bai, Yechao¹  Chen, Yang¹  Tang, Lan^1,2  Liu, Xin³ 

[1] Nanjing Univ, Sch Elect Sci & Engn, Nanjing 210023, Jiangsu, Peoples R China

[2] Southeast Univ, Natl Mobile Commun Res Lab, Nanjing 210023, Jiangsu, Peoples R China

[3] Univ Oulu, Ctr Machine Vis & Signal Anal, Oulu 90014, Finland

关键词: Image denoising;    Group sparsity residual constraint;    Nonlocal self-similarity;    Gaussian Mixture Model;    Iterative shrinkage algorithm;   

DOI  :  10.1016/j.neucom.2017.11.004

来源: Elsevier

【 摘 要 】

Nonlocal image representation has been successfully used in many image-related inverse problems including denoising, deblurring and deblocking. However, most existing methods only consider the nonlocal self-similarity (NSS) prior of degraded observation image, and few methods use the NSS prior from natural images. In this paper we propose a novel method for image denoising via group sparsity residual constraint with external NSS prior (GSRC-ENSS). Different from the previous NSS prior-based denoising methods, two kinds of NSS prior (e.g., NSS priors of noisy image and natural images) are used for image denoising. In particular, to enhance the performance of image denoising, the group sparsity residual is proposed, and thus the problem of image denoising is translated into reducing the group sparsity residual. Because the groups contain a large amount of NSS information of natural images, to reduce the group sparsity residual, we obtain a good estimation of the group sparse coefficients of the original image by the external NSS prior based on Gaussian Mixture Model (GMM) learning, and the group sparse coefficients of noisy image are used to approximate the estimation. To combine these two NSS priors better, an effective iterative shrinkage algorithm is developed to solve the proposed GSRC-ENSS model. Experimental results demonstrate that the proposed GSRC-ENSS not only outperforms several state-of-the-art methods, but also delivers the best qualitative denoising results with finer details and less ringing artifacts. (c) 2017 Elsevier B.V. All rights reserved.

【 授权许可】

Free