Performance Comparison of ROAD Statistic Based Nonlinear Filters for Image Denoising

Removal of impulse noise from images without losing their features is an important requirement in image processing applications. The standard median filter is a good candidate for eliminating impulse noise; however, it fails to preserve the image details when the window size is made larger. In this paper, three nonlinear filtering schemes, based on the recently reported rank-ordered absolute difference (ROAD) image statistic, are introduced for cleaning the images confounded by impulse noise. These ROAD based nonlinear filtering techniques perform the filtering function on the basis of sum of rank-ordered absolute differences between central pixel inside the window and its neighboring pixels. Extensive simulation studies have been carried out by applying the proposed filtering schemes on a test image corrupted with different levels of impulse noise and the results obtained are included.