Macroblock based bit allocation for snr scalable video coding with hierarchical B pictures

We investigate SNR scalable video coding based on motion compensated temporal prediction with hierarchical B pictures. This structure is a fundamental part of the scalable extension of H.264/AVC (SVC), which is currently under standardization. Due to SNR scalability, reconstruction of inter blocks may cause error accumulation within the B picture hierarchy, which is called the drift effect. In this paper we investigate the error propagation considering coding control, motion information, and target quality. We first consider open loop coding control and generalize to the case of one or multiple loops. We present a practical algorithm to quantify locally varying drift properties, which is utilized for performing rate-distortion optimized macroblock based bit allocation. We compare the approach to picture based bit allocation within the SVC test model and observe coding gains up to 0.4 dB.