Reduction of coding artifacts using chrominance and luminance spatial analysis

摘要

This paper presents a new low-cost method for coding artifact reduction of MPEG compressed video sequences. The algorithm detects possible locations of blocking and ringing artifacts by analyzing local spatial activity of luminance and chrominance components. The analysis explores differences in processing of luminance and chrominance during video coding. Depending on the outcomes of the analysis, one of three low-pass filters is applied. The results of our experiments show the high efficiency of the proposed approach.