基于压缩感知的视频编码技术研究

通过分析和比较基于传统视频编码架构和基于分布式视频压缩感知的视频压缩感知编码技术的主要特性,提出一种新的基于帧间模式决策的混合式视频压缩感知编码模型.针对不同的图像内容变化速率、感知帧测量率以及图像组尺寸等情况,分别进行算法的仿真实验.测试结果表明,所提出的混合式视频压缩感知编码技术能够在高保真重建视频序列的同时,兼顾编解码速度.

Research on the video coding technology based on compressive sensing

Through analyzing and comparing principal characteristic of video compressive sensing coding technology based on the structures of traditional video encoding and distributional compressive video sensing, a new kind of blending compressive sensing video model based on mode judging between frames is proposed. The simulating experiments for this algorithm are carried on respectively, referring to variable rate of image content, measure rate of CS frame and size of image group from different video sequences. The test result indicates that the proposed coding technology based on blending video compressive sensing can not only reconstruct video sequence with high fidelity, but also pay proper attention to encoding and decoding speed.