Rate-distortion optimization of macroblock-based progressive fine granularity scalable video codec

This paper proposes a framework for a rate-distortion optimal progressive fine granularity scalable (PFGS) video codec. Firstly, we introduce the use of multiple enhancement layer references and a modified mode decision algorithm based on the analysis of the unique characteristic of residue data in the enhancement layer. Secondly, the rate-distortion optimal method to select the enhancement layers is discussed and a fast algorithm is proposed to solve the rate-distortion optimization problem by tracing the convex hull of the rate-distortion (R-D) curve. Experimental results show that coding efficiency can be improved by 0.5 dB on average in PSNR comparing to that of the current PFGS codec.