J.K. Mandal and Aparna Sarkar, University of Kalyani, India

In this paper a Modified Directional Weighted Cascaded-Mask Median (MDWCMM) filter has been proposed, which is based on three different sized cascaded filtering windows. The differences between the current pixel and its neighbors aligned with four main directions. A direction index is used for each edge aligned with a given direction. Then, the minimum of these four direction indexes is used for impulse detection for each and every masking window. Depending on the minimum direction indexes among the three windows one window is selected. The filtering is done on this selected window. Extensive simulations showed that the MDWCMM filter provides good performances of suppressing impulse with low noise level as well as for highly corrupted images from both gray level and colored benchmarked images.

Modified Directional Weighted Cascaded-Mask Median filter (MDWCMMF, PSNR, SNR, Median, gray & color image.