降低帧率中的运动矢量合成研究  

The Motion Vector Composition Research in Reducing Temporal Resolution

作　　者：董武[1] 

机构地区：[1]北京印刷学院信息与机电工程学院,北京102600

出　　处：《北京印刷学院学报》2007年第6期32-34,共3页Journal of Beijing Institute of Graphic Communication

基　　金：北京印刷学院青年基金资助项目(E-b-07-43)

摘　　要：在异构网络之间进行视频的码率转换中,为了降低视频的编码量,基于降低时间分辨率模型的转换是一种重要的方法。在DCT域中分析了视频邻近帧中物体的运动情况,为了充分利用源码流中的运动矢量,提出了一种新的运动矢量合成方法:加权综合(weighted integration,WI)算法。仿真实验结果表明,WI算法和线性内插法、FDVS算法相比,提高了视频的转换质量。Transcoding, based on the reduced temporal resolution model, is an important method to reduce code quantity in a heterogeneous network. The object＇s motion in the near frames is analyzed in DCT field, and a new motion vector composed method, weighted integration algorithm, WI, is proposed in order to make the best use of source video＇s information. The experiment result reveals that WI algorithm improves the transcoding quality of the video, compared with the bilinear interpolation of algorithm and FDVS algorithm.

关 键 词：视频 码率转换 降低帧率 合成运动矢量 

分 类 号：TN919.85[电子电信—通信与信息系统]