The Study, Based On The H.264/avc Video Transmission Technology

Based on the analysis of the features and error concealment tech-niques of H.264, this dissertation optimizes H.264 by reducing its com-putational complexity, refining its rate distortion model and increasing its reliability.From the viewpoint of real-time encoding, the optimization for time-consuming modules is necessary to reduce the computational com-plexity to meet the requirements of real-time applications, because the improved performance of H.264 is at cost of increasing the algorithms'computational complexity. Based on the study of the intra-mode selection algorithm of H.264, an edge oriented fast intra-mode decision algorithm is proposed to reduce the computational complexity.The accurate statistical distribution model for the integer transform coefficients is very important for the improvement of coding efficiency. Based on the study of the difference between the distribution of transform coefficients and the classical Laplace distribution, a separate Cauchy rate distortion model is proposed, in which separate distributions are em-ployed for luma and chroma components. The new model better reflects the statistical distribution of the transform coefficients.Due to the congestion and noise interference, packets lost and trans-mission error are inevitable. Since compressed video data are very sensi-tive to interference and packet lost, error resilience and error concealment technique become very important in video transmission. Based on the study in existing error concealment technologies, some improved error concealment algorithms for H.264 are proposed by using spatial correla-tion, temporal correlation, and mixture correlation in the video seuences are proposed. These algorithms improve the robustness of coding systems, and have some application value.