Temporal concealment of packet-loss related distortions in video based on Structural Alignment

Concealment of packet-loss related distortions in multimedia communications assumes great importance with the emergence of packet switched communication using the IP protocol. This paper proposes a novel temporal algorithm to conceal packet-loss related distortion in video transmission. It is based on the insight that the spatial region around the area in the previous frame that yields the best concealment will in general be highly correlated with the corresponding area around the region to be concealed in the present frame. The proposed structural alignment algorithm (SAA) consistently yields better concealment and exhibits a superior ability to align edges. We compare performance of the proposed algorithm with that of a well known side matching algorithm (SMA). Finally, we note the strengths and the weaknesses of SAA and SMA and propose a combined concealment algorithm (CCA) that uses the strengths of both, SAA and SMA for more effective concealment. With their superior concealment ability, the SAA and CAA can find application in real-time concealment of packet-loss related artifacts in streaming video.