Motion compensated ASMKF for restoration of color video sequences corrupted by impulsive noise

Various nonlinear filters have been proposed in the literature for restoration of video sequences contaminated by impulsive noise. Among these, median filters and their variants are the most popular filter types used for impulsive noise suppression. However, median filters and their variants tend to remove fine image details and destroy fine texture, especially when the signal to noise ratio is low. In order to circumvent the limitations of the median filters, an adaptive simplified-model Kalman filter (ASMKF) is proposed for restoration of video sequences corrupted by impulsive noise. For motion estimation, a block matching technique is used. The efficacy of the proposed filter is illustrated through implementation results on the restoration of a color video sequence.