Gabor Feature Based Nonlocal Means Filter for Textured Image Denoising

Abstract

Image denoising is still a challenging one as day to day more and more techniques and implementation comes. Moreover, whatever the technique we use, it has its own set of merits and demerits. The nonlocal means (NLM) filter has its own rewards that beat the conventional signal denoising algorithms. Non Local Means filter recovers noise-corrupted images by relocating the pixel value with the weighted pixel, where each weight is termed depending upon the Gabor-based texture measure. The purpose of the Gabor feature based Non-Local Means filter for textured signal denoising. The conventional Non Local Means filter and four other signal denoising techniques in textured images are dissipated by Additive Gaussian Noise (AGN). Our outcomes show that the proposed filter can denoise textured signals more effectively.