A Self-Adaptive Restoration Algorithm for Image Corrupted with Strong Noise Based on Group Sparsity Residual Constraint

作者：Gao, Hongxia; Chen, Zhanhong; Zeng, Runhao; Luo, Lan; Chen, An^*; Ma, Ge

来源：Journal of South China University of Technology(Natural Science Edition), 2018, 46(8): 11-18.

DOI：10.3969/j.issn.1000-565X.2018.08.002

• Adaptive regularization
• Group sparsities
• Images corrupted with strong noise
• Patch matching
• Self-similarities

Compressed sensing based on group sparsity has shown great potential in image denoising. However, most existing methods considered nonlocal self-similarity prior of noisy images only in a block-wise manner, which reduced reconstruction quality. This paper introduces group sparsity residual and total variance as the regularization constraints into image restoration model based on group sparsity and transforms the reconstruction problem into multiscale patch matching and decreasing group sparsity residual. Then, a self-adaptive iterative algorithm for image restoration is proposed based on estimating original images'group sparse coefficients and matching patches at multiple scales, which improve group sparsity learning's performance in denoising and restoring fine structure. Experimental results demonstrate that the proposed algorithm can retain more details, reduce over-smoothness and artifacts. The proposed algorithm outperforms the contrast benchmarking algorithms for images corrupted with strong noise, such as BM3D, WNNM when considering the visual results and the objective evaluation together.

• 出版日期2018-8-1
• 单位华南理工大学