Fast Mode Decision Algorithm for Spatial and Coarse Grain Quality Scalable Video Coding

In the current scalable video coding (SVC) extension of the H.264/AVC standard, an exhaustive search method is employed to achieve the optimal rate-distortion (RD) cost, but it results in an extremely heavy computational burden. In this paper, an improved fast mode decision algorithm is proposed for spatial and coarse grain quality (CGS) scalable video coding. It makes use of the mode correlation between the base layer and the enhancement layer as well as the mode correlations between macroblocks (MBs) in spatial neighborhood. Specifically, the prediction mode candidates include the mode predicted from the co-located MB in the base layer, the best modes of spatial neighboring MBs, and the SKIP mode. Combined with an early termination check and a refinement search, the coding complexity can be reduced while preserving good quality. From the experimental results, the proposed method can reduce the total encoding time by up to 52.25% with negligible quality degradation.