基于高效视频编码的运动估计算法综述  

Review of motion estimation algorithm for high efficiency video coding

作　　者：但鸿键 汪伟[1] Dan Hongjian;Wang Wei(School of Optoelectronic Information and Computer Engineering,University of Shanghai for Science and Technology,Shanghai 200093,China)

机构地区：[1]上海理工大学光电信息与计算机工程学院

出　　处：《计算机时代》2020年第1期1-4,9,共5页Computer Era

基　　金：上海市教委2016年青年教师培养资助计划项目（ZZs115012）;上海理工大学光电学院教师创新能力建设项目（1000302006）资助

摘　　要：运动估计是高效视频编码标准中的重要技术,通过帧间的运动矢量估计有效地减少视频中的时间冗余,进而提高视频编码的压缩效率。文章分析了运动估计算法的基本原理,将算法划分为空域和频域两大类,并进一步将空域类算法细分为块匹配法、光流法、像素递归法和贝叶斯法等,将频域类算法分为相位法、离散余弦变换法和小波域法,详尽地讨论了在高效视频编码应用过程中各种方法的优缺点。最后,对基于高效视频编码的运动估计算法进行总结,探讨了发展趋势。Motion estimation is an important technology in high efficiency video coding standard. Motion vector estimation between frames can effectively reduce the time redundancy in video and improve the compression efficiency of video coding. This paper analyzes the basic principle of motion estimation algorithm, divides the algorithm into two categories, i. e. spatial domain and frequency domain, and further subdivides the spatial domain algorithm into block matching method, optical flow method, pixel recursive method and Bayesian method, and divides the frequency domain algorithm into phase method, discrete cosine transform method and wavelet domain method. The advantages and disadvantages of the application of various methods in efficient video coding are discusses in detail. Finally, the motion estimation algorithms for high efficiency video coding are summarized, and the corresponding development trends are discussed.

关 键 词：视频编码 空域 频域 运动估计算法 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]