基于运动矢量可分级的视频编码方法  

Method of Video Coding Based on Motion Vector Scalability

作　　者：袁嘉晟[1] 方志军[1] 叶金财[1] 

机构地区：[1]江西财经大学信息管理学院,江西南昌330013

出　　处：《电视技术》2010年第4期19-21,35,共4页Video Engineering

基　　金：国家自然科学基金项目(60862002);江西省青年科学家(井冈之星)培养对象计划资助项目

摘　　要：传统的视频可分级编码,未考虑到运动矢量编码的渐进传输。特别是在低码率情况下,运动矢量占据了视频的大部分码流,其中有些运动矢量幅值较小,对其后的运动估计和运动补偿预测的影响并不大,这些运动矢量所占用的比特流还是一定的,由于未能采取可分级技术,导致编码效率降低。因此,提出一种基于运动矢量可分级的视频编码方法,对运动矢量模进行判断,截取其比特流,以达到质量可分级,并通过实验分析,说明该方法的可行性。The traditional video scalable coding does not consider progressive transmission of motion vector coding. Especially in the case of low bit-rate motion vectors occupy large part of video bitstream. However some motion vectors＇ magnitudes are lower and these vectors not only have little or no effect on following motion estimation and motion compensation prediction, but also still occupy some certain part of bitstream. No taking scalable technology leads to the reduction of coding efficiency. And thus, a method of video code is proposed based on motion vector scalability which is to intercept motion vector bitstream after judging norm of motion vector to achieve quality scalability. As well as it aims to validate the feasibility of this method by analyzing experiment.

关 键 词：运动矢量 可分级视频编码 运动估计 

分 类 号：TN919.81[电子电信—通信与信息系统]