A rate-distortion optimized real time intra-update method for packet video transmission

Error resilience to packet loss is a critical requirement in predictive video coding for transmission over packet-switched networks, since the prediction loop propagates error and causes substantial degradation in video quality. This paper proposes an algorithm based on the rate-distortion optimization model of H.263 to approximately estimate the quantization distortion of one frame. By analyzing the overall distortion of the decoder, an algorithm to estimate the rough propagation is presented. Based on these acquirements, a real time intra-update criterion is proposed, which reduces the computational complexity significantly. Through comparing the increased quantization distortion by intra-updating and reduced need to intra-update a macroblock, the encoder can quickly decide whether there is a need to intra-update a macroblock. Based on TMN8 coder, the performance of the proposed algorithm and the traditional R-D optimized intra-update method is compared. Simulation results demonstrate that proposed method achieves the same or even better gains in PSNR over traditional method.