摘要
测量矩阵是压缩感知编码端不可或缺的一部分,测量矩阵的好坏直接影响到重构出信号的质量好坏。目前常用的测量矩阵有很多,但很少有文献将该矩阵进行系统地比较。对目前常用的测量矩阵做综合分析,比较它们在压缩感知编解码过程中重构出的信号在视觉效果、峰值信噪比、重构时间等方面的区别。另外,由于使用分块的方式对图像进行编解码,因此也分析比较了不同的分块大小对重构的影响。
Measurement matrix is an integral part of the encoding side in compressive sensing. It has a direct impact on the reconstructed signal＇ s qual- ity. There are many kinds of measurement matrices now. But there is few literature which does a comparative analysis of these matrices. The purpose of this paper is to do a comprehensive analysis on visual effect, peak signal-to-noise ratio, reconstruction time of these commonly used measurement matri- ces in compressive sensing. In addition, since the picture is encoded based on block, so different block size＇ s impact on the reconstruction is also ana- lyzed and compared in this article.
出处
《电视技术》
北大核心
2014年第5期1-4,共4页
Video Engineering
基金
华为公司创新研究计划资助项目
关键词
压缩感知
测量矩阵
重构
compressive sensing
measurement matrix
reconstruction
