Cascaded decision based Non Linear filters for high density outlier noise removal in images and videos

This work aims a cascaded decision based filter for the removal of high density in image and video is proposed. The proposed scheme operates in two levels. First level is a decision based median filter and the later employs a decision based unsymmetrical trimmed variants. In the first stage the corrupted pixels are found and replaced by the median of the current processing window. The second level again inspects for the occurrence of outliers and eliminates them using unsymmetrical trimmed variants depending upon the content of the current processing window. The Main idea of the proposed algorithm is to formulate an algorithm that eliminates the outliers for high density outlier noise. The proposed algorithm is found to exhibit excellent noise elimination capability for noise densities as high as 90%. The proposed scheme shows better results quantitatively and qualitatively in image and video than the standard and other cascaded algorithms.