Compressed image error concealment and post-processing for digital video recording

The recently developed MPEG video compression standards can compress motion pictures very efficiently; however, the compressed data are rather sensitive to channel noise. We therefore propose a robust decoder that can tolerate errors. Particularly, a two-priority coding scheme is suggested to increase the overall performance. This priority assignment can also be used to implement the trick playback mode for a DVCR system. Moreover, an error concealment algorithm is proposed to compensate the cell loss effect due to random noise. Finally, the image artifacts due to undetected errors and/or imperfect error concealment can be reduced by a properly designed post-processing scheme. As shown by simulations, damaged video sequences can be recovered almost free from visible artifacts. Our study is different from previous approaches in that we design a complete decoder system and the system is simulated in an environment that is close to reality.