Title :
Structural similar patches for nonlocal-means with modified robust M-estimator and residual images
Author :
Xiong, Bo ; Yin, Zhouping
Author_Institution :
State Key Lab. of Digital Manuf. Equip. & Technol., Huazhong Univ. of Sci. & Technol., Wuhan, China
Abstract :
Nonlocal-means provides a powerful framework for denoising. However, there are some issues of it. In this letters, we proposed some improvements for solve or alleviating the influence of the issues. First, we show the connection between the NL-means and robust M-estimator framework, and propose a modified robust M-estimator function for NL-means. Then, we introduced the structure similar for searching the similar patches. And then, we propose a iterative theme for using it to improve the quality. Finally, extensive simulation results reveal that the proposed improvements are useful for denoising procedure.
Keywords :
estimation theory; image denoising; iterative methods; NL-means; image denoising; iterative theme; nonlocal-means; residual image; robust M-estimator; structural similar patches; Image denoising; Noise measurement; Noise reduction; PSNR; Robustness; Simulation; Image denoising; Nonlocal-means; residual image; robust M-estimator; structural similarity;
Conference_Titel :
Mechatronics and Automation (ICMA), 2011 International Conference on
Conference_Location :
Beijing
Print_ISBN :
978-1-4244-8113-2
DOI :
10.1109/ICMA.2011.5985748