Rate Control Based Fast Intra-Update Algorithm

Predictive video coding for transmission over packet-switched networks is very sensitive to channel errors, since the prediction loop propagates errors in both the spatial and the temporal directions, and causes substantial degradation in video quality. Consequently, research on enhancing the error resilience of video transmission has been widely addressed.This paper is mainly concerned with Intra-update. First, the principle of video coding and main compression technologies are introduced, and the error propagation problem and corresponding error control methods are analyzed. In succession, the advantages and disadvantages of common Intra-update arithmetic are pointed, and the rate-distortion framework based Intra-update method is referred as a perfect solution. However, based on Lagrange equation, classical Intra-update methods need to do actual encoding and decoding operations to calculate the rate and distortion, so the computational complexity is very huge, so that it is not suitable to some real-time applications. By analyzing the rate-distortion model and rate control method in TMN8, this paper proposes a rate control based fast Intra-update method. Through the end-to-end distortion analysis, the total distortion is suggested to be divided into quantization distortion, propagated distortion and concealment distortion to be calculated individually. A distortion modal based on rate-distortion modal in TMN8 is proposed to approximately estimate the total quantization distortion when the available rate is prescribed. Through comparing the increased quantization distortion and reduced propagation distortion, the encoder can quickly decide whether need to Intra-update a macroblock without actual encoding and decoding operations. Thus, it reduces the computational complexity significantly. Simulation results demonstrate that the proposed method achieves the same or even better gains over traditional method.