Blocking effect removal using regularization and dithering

Image compression codecs suffer from the various coding artifacts at low bit rate. The coding artifacts can be removed by way of post-processing. Most of the coding artifacts removal algorithms depend on a priori knowledge of the original image, which is about its smoothness. The resulting surface of the restored image is often too smooth to be the one in the natural scene. Dithering is based on the idea of adding controlled noise to the system to achieve the better results. An algorithm based on regularization and dithering is proposed to remove the coding artifacts. The results show that the algorithm removes the coding artifacts successfully and the surface of the restored image is more visually pleasing.