针对H 2 6 4的多模式运动估计 ,引入运动无序度的概念来刻画当前宏块的运动混乱程度 ,通过理论和实验统计分析 ,建立了无序度和最优块模式之间的关联模型 基于该模型 ,提出了一种基于当前宏块运动场无序度的块模式判定算法 (FastModeSelectionalgorithmbasedonChaosDegreeofMotionField ,CDMF FMS) ,最后的实验结果表明 ,该算法在基本保持原来参考软件编码性能的基础上 ,整体编码时间平均减少 2 0 %以上 Aiming at the multi-modal motion estimation of H 2 6 4, the concept of motion disorder degree is introduced to describe the motion disorder of the current macro-block. Through the theoretical and experimental statistical analysis, the correlation model between the disorder degree and the optimal block mode is established Based on this model, this paper proposes a FastModeSelectionalgorithmbasedonChaosDegreeofMotionField (CDMFFMS) based on the degree of disorder of the current macroblocks. The final experimental results show that the proposed algorithm, while basically keeping the original reference software coding performance, The average time to reduce 20% or more