New Diffusion Based Image Denoising Techniques; Evaluations and Comparisons

Abstract

This paper introduces three new diffusion based methods for image denoising in which three new neighboring structure are used. These structures are aimed to consider the gradual nature of edges, which are preserved in the diffusion based methods. Different experiments with different conditions were done to evaluate the methods and also compare with the two well known denoising algorithms: wiener and median filter. For reasonable evaluation and comparison, the methods are compared using different performance measures; the results express the considerable superiority of the proposed methods.