Removal of Impulse Noise using Iterative Unsymmetrical Trimmed Median Filter

Abstract

is universally accepted that Median filter is the best filter known so far. Based on this fact many variants of median filter were developed to improve the performance of the standard median filter. In this paper a new approach for the restoration of gray scale and color images that are highly corrupted by impulse noise is proposed. The algorithm works on low density noise also. The algorithm has three stages - firstly, finding the corrupted pixels, secondly de-noising the corrupted pixels; thirdly, minimizing the de-noised image to root image. The article proves that the new approach is guaranteed to converge to root image within a finite number of iterations. The proposed algorithm shows better results than the Standard Median Filter, Recursive Median Filter and Decision based Unsymmetrical Trimmed Median Filter.