WCSE

Steering Kernel Regression Total Variation for Image Denosing

Abstract— In this paper, a novel l1 minimization model was proposed for image denosing. We firstly proposed a new regularization term called Steering Kernel Regression Total Variation(SKRTV), which exploits the local structural regularity properties in natural images. By combining the SKRTV regularization term and global fidelity term, we proposed a maximum a posteriori probability framework of image denoising. Furthermore, split Bregman iteration was applied to implement the proposed model. Extensive experiments demonstrated the effectiveness of the proposed method.

Index Terms— steering kernel regression, total variation, structural regularity, split Bregman iteration.

Lin Li, Xinhua Wei, Zhiyu Zuo, Wenjing Zhu
School of Agricultural Equipment Engineering, Jiangsu University, CHINA

[Download]

Cite: Lin Li, Xinhua Wei, Zhiyu Zuo, Wenjing Zhu, "Steering Kernel Regression Total Variation for Image Denosing," Proceedings of 2016 6th International Workshop on Computer Science and Engineering, pp. 788-791, Tokyo, 17-19 June, 2016.