基于H．264标准的运动估计参数性能分析

H．264是新一代的视频编码标准，在相同图像质量下，编码比特率比MPEG-2、H．263HLP和MPEG-4ASP分别减少64．46％、48．80％和38．62％，但是H．264编码性能的提升是以计算复杂度的大幅度增加为代价的．在介绍多预测块模式搜索及率失真优化模式选择这2种带来高编码效率的技术的基础上，通过实验研究来分析这2种技术对编码性能的影响．结果表明，这两项技术一般都只能带来编码性能的微略提高，而编码时间即计算复杂度却显著增加．从而为H．264标准在实际视频编码工程应用中的参数选择提供一个重要参考．

Performance Analysis of Motion Estimation Parameters Based on H. 264

H, 264 is the new standard for video coding, it can save 64.46%, 48, 80% and 38.62% bit-rate compared with that of MPEG-2, H. 263 HLP and MPEG-4 ASP respectively under the same decoded picture quality, But the high encoding efficiency is at the cost of heavily increased computational complexity. The multi-prediction macroblock mode search and rate-distortion optimization mode decision that achieve high encoding efficiency are introduced firstly. Then experiment researches are performed to analyze their effect on encoding efficiency. Results show that these two kinds of technique can only achieve slightly improvement on encoding efficiency, but computational complexity increases heavily. This provides an important reference for the parameters selection when H. 264 is used in the practical application of video coding.