On the adaptive impulsive noise reduction in color images

In this paper a novel class of noise attenuating and edge enhancing filters for color image processing is introduced and analyzed. The proposed adaptive filter design is minimizing the cumulative dissimilarity measure of a cluster of pixels belonging to the sliding filtering window and outputs the centrally located pixel. The proposed filter is computationally efficient, easy to implement and very effective in suppressing impulsive noise, while preserving image details and enhancing its edges. Therefore it can be used in any application in which simultaneous denoising and edge enhancement is a prerequisite for further steps of the color image processing pipeline.