一种基于LDPC矩阵的压缩感知测量矩阵的构造方法.pdfVIP

第 37 卷第3 期 电子测量技术 -11制崎晦11 ELECTRONIC MEASUREMENT TECHNOLOGY 2014 年 3 月 一种基于 LDPC 矩阵的压缩感知 测量矩阵的构造方法 II1c)j1神刘里张淑* 〈天津大学电子信息工程学院 天津 300072) 摘 要:在一些对采样数据速率有严格要求的实际应用中，对低采样率的压缩感知具有广泛需求。基于 LDPC 矩阵 的特点，提出了一种类似托普利兹矩阵的压缩感知测量矩阵，所提出的测量矩阵构成方法易于实现。仿真结果表明， 在低采样率的情况下，采用本文所提方法构造的测量矩阵不仅有着与常用稀疏测量矩阵相比更好的稀疏性，且将其用 于图像压缩感知时可获得较好的图像重构质量。 关键词:压缩感知;测量矩阵z 托普利兹矩阵;低采样率;图像压缩 中图分类号: TP751 文献标识码:A 国家标准学科分类代码: 580.80 Coostructioo method of CS measuremeot matrices based 00 LDPC matrices Tian Peipei Liu Yu Zhang Shufang (Electronic and Information Engineering. Tianjin University. Tianjin 300072.China) Abstract: In some practical applications having stringent requirement for the sampling rate. compressed sensing of low sampling rate is needed widely. In this paper. we proposed a compressed sensing measurement matrix based on the characteristics of LDPC matrices. which is similar to Toeplitz measurement matrix. The proposed method of constructing the matrix is easy to implement. The simulation results show that in the case of low sampling rates. the constructed measurement matrix using the method proposed in this paper has better sparsity compared with conventional sparse measurement matrices.and which can obtain better image reconstruction quality when used in image compressed sensing. Keywords: compressed sensing; measurement matrix; Toeplitz matrices; low sampling rate; image compression 压缩感知的测量矩阵，如稀疏随机测量矩阵[町、LDPC 矩 1 引言