Simulated evaluation of new switching based median filter for suppressing SPN and RVIN

Abstract

In the past two decades, the SPN (salt and pepper noise) suppressing method is worldwide interested researches on computer vision and image processing hence many SPN suppressing methods have been proposed. In general, the primary goal of SPN removal method is the suppressing of SPN in digital images thereby one of the recent effective and powerful SPN suppressing methods is a new switching-based median filtering (NSMF), which is innovated for suppressing high density SPN. Consequently, this paper thoroughly examines its efficiency and constrain of a new switching-based median filtering when this filter is used for contaminated image, which is synthesized by SPN and RVIN (random-value impulsive noise). In these simulations, six well-known images (Lena, Mobile, Pepper, Pentagon, Girl, Resolution) with two impulsive noise classes (SPN and RVIN) are used for measuring the its efficiency and constrain. An evaluation of the efficiency is conducted with many previous methods in forms of subjective and objective indicators.