A real-time pattern selection algorithm for very low bit-rate video coding using relevance and similarity metrics

Very low bit-rate video coding using regularly shapedpatterns to represent moving regions in macroblocks has good potentialfor improved coding efficiency. This paper presents a realtimepattern selection (RTPS) algorithm, which uses a pattern relevanceand similarity metric to achieve faster pattern selection froma large codebook. For each applicable macroblock, the relevancemetric is applied to create a customized pattern codebook (CPC)from which the best pattern is selected using the similarity metric.The CPC size is adapted to facilitate real-time selection. Resultsprove the quantitative and perceptual performance of RTPS is superiorto both the Fixed-8 algorithm [16] and H.263.