Video Mosaic Block Detection Based on Template Matching and SVM

Mosaic block is easily occurred in the TV signals, which will cause the degraded video. In this paper, we propose an automatic degraded video detection approach based on support vector machine and template matching to detecting mosaic block in video. We develop a detection algorithm based on a cascade of processing steps as follows. First, we use DWT to remove the vertical high frequency of a frame and use the Canny edge algorithm to detect the frame. Second, we use the template matching to find the candidate mosaic block coordinate. Third, a SVM classifier is employed to distinguish mosaic block from the candidate ones. When the number of mosaic block is higher than a threshold, we can regard the frame as the frame with mosaic block. Experimental results show that the proposed algorithm has good performance.