摘要
压缩感知理论突破传统采样定理的限制,为研究信号处理提供了新思路,极大地吸引了相关研究人员的关注。将采用稀疏基为小波基、高斯随机矩阵为观测矩阵,贪婪类算法和基于l2范数的迭代重加权最小二乘算法(IRLS)作为图像重构算法,并进行了仿真与比较。经过仿真分析,IRLS算法重构的峰值信噪比较大,重构质量较好,但时间较长。在对重构质量要求不是很高时,可选用贪婪类算法中运行速度较快的ROMP算法。
Compressed sensing theory breaks through the limitations of the traditionalsampling theorem and provides a new way of thinking for the study of signal processing.Since it was put forward,it has attracted the attention of relevant researchers.In this pap-er,wavelet basis is used as the image sparse matrix,Gauss random matrix as observationmatrix,greedy algorithm and norm-based iterative weighted least squares algorithm(IRLS)areused as image reconstruction algorithmsand simulation and comparison are made.It isfound that,the peak signal-to-noise ratio(PSNR)reconstructed by IRLS algorithm is larger,the reconstructed quality is better,while the time is lo-nger after simulation analysis.When the quality of reconfiguration is not very high,the g-reedy ROMP algorithm which runs faster could be chosen.
作者
张珊珊
赵建华
Zhang Shanshan;Zhao Jianhua(Xi′an Technological University,Electronic Information Engineering,Xi′an 710021,China)
出处
《国外电子测量技术》
2019年第10期44-48,共5页
Foreign Electronic Measurement Technology
关键词
压缩感知
重构算法
IRLS算法
贪婪算法
compressed sensing
reconstruction algorithm
IRLS algorithm
greedy algorithm
