Removal of high density impulse noise from color images using an adaptive fuzzy filter

Abstract

Low density impulse noise can be successfully denoised by traditional median filter. But the performance badly deteriorates if the noise density is high. Recently fuzzy-based adaptive median filter has been introduced to handle high density noise in grayscale images. To make it more generalized in this paper we develop an adaptive fuzzy median filter to remove impulse noise from color images. Here we use fuzzy filter with fully adaptive and automatically adjustable threshold value on each component of RGB color space to remove the noise. Finally concatenation is applied to RGB components for restoring the original image. Experimental result confirms the superiority of the proposed method compared to the state-of-the-art median filtering approaches in terms of both denoising and detail preservation.