Image restoration with l2-type edge-continuous overlapping group sparsity

• Published in: Pattern recognition letters Vol. 68; pp. 211 - 216
• Main Authors: He, Xiaowei, Fan, Junli, Zheng, Zhonglong
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 15-12-2015
• Subjects: Edge-continuous; Image restoration; Inverse problem; Overlapping group sparsity; Image restoration; Overlapping group sparsity; Inverse problem; Edge-continuous
• ISSN: 0167-8655; 1872-7344
• DOI: 10.1016/j.patrec.2015.10.001

•We propose a novel regularizer named EC-OGS for image processing.•A new model based on the EC-OGS regularizer is proposed.•We propose a new method for computing the proximal solution of the key subproblem.
It is important and necessary to take account of the non-zero pattern in image sparsity representation. In this paper, we present an image restoration model by introducing a novel edge-continuous overlapping group sparsity regularizer (EC-OGS), based on our observation that the non-zero entries in an image gradient domain often distribute along its edges. The model is solved by the ADMM (alternating direction method of multipliers), where a fast novel algorithm is proposed for computing the proximal operator in solving the subproblem with EC-OGS regularizer. The proposed model can be applied to various image restoration tasks including denoising, deblurring, and edge-detecting. The numerical experiments demonstrate the effectiveness of our method in terms of PSNR, visual effect and edge preserving.