First order neighborhood decision based median filter

In this paper, a first order neighborhood decision based median filter algorithm for restoration of gray scale images that are highly corrupted by impulse noise is proposed. The proposed algorithm detects the noisy pixels and considers the first order neighborhood pixels of the detected noisy pixel for further decision. If all the first order pixels are 0's and 255's, then the noisy pixel is replaced by the average values of these first order pixels. When all the pixels in the first order neighborhood are not 0's and 255's, the noisy pixel is replaced by trimmed median value of the first order neighboring pixels. The proposed algorithm shows better performance than the standard median filter, peer group based filter and Modified Decision Based Unsymmetric Trimmed Median Filter.