Compression artifact reduction based on total variation regularization method for MPEG-2

In accordance with the recent improvement in the quality of image displays for consumer electronics, digital image compression artifacts are visible than ever. Noise removable methods are key issues not only in research areas such as image sensing and satellite imaging but also in consumer electronics such as television signal conversion, DVD and Blu-ray encodings. Among various compression artifact removable methods, the Total Variation (TV) regularization method is one of the most promising solutions. However, there are two problems: one is loss of image minuteness in the texture components, and the other is insufficient compression artifact removal in low bit rates. In this paper, we propose a new noise removable method utilizing the TV regularization method for moving pictures compressed by MPEG-2. We obtain fine experimental results in image quality improvement, and our method is possible to be applied to the other DCT based-image compression standards such as MPEG-4 and H.264.