Rejection of scratches from video by use of filters with prediction error processing

This paper deals with denoising of color video sequences using improved median-based filters appropriate for realtime implementations. In the paper, it is proven that application of prediction error processing results in improved efficiency of video restoration. The basic idea is to predict a pixel value by using a nonlinear filter, and then to calculate the prediction error by a comparison with the input value that maybe is corrupted. The output value is a sum of the prediction and the prediction error processed according to a nonlinear function. The experimental results prove that such filters are appropriate for rejection of scratches from video even in the interlaced format. The computational overhead is less than 10% as compared to classic median-based filters.