首页 | 官方网站   微博 | 高级检索  
     

迭代硬阈值压缩感知重构算法——IIHT
引用本文:张宗念,李金徽,黄仁泰.迭代硬阈值压缩感知重构算法——IIHT[J].计算机应用,2011,31(8):2123-2125.
作者姓名:张宗念  李金徽  黄仁泰
作者单位:1. 东莞理工学院 电子工程学院,广东 东莞523808 2. 东莞理工学院 网络中心,广东 东莞523808 3. 东莞理工学院 计算机学院,广东 东莞523808
基金项目:广东省自然科学基金资助项目
摘    要:研究了压缩感知信号重构算法的理论,针对迭代硬阈值(IHT)重构算法对测量矩阵的过分依赖、计算复杂度高、运算时间长的缺点,通过修订迭代硬阈值重构算法的代价函数和自适应地调整迭代步长的选取原则,设计了一种迭代硬阈值重构算法--IIHT。IIHT算法显著提高了信号精确重构的概率,降低了算法的计算复杂度,进一步减少了算法的运算时间,加快了算法的收敛速度。

关 键 词:迭代    硬阈值    压缩感知
收稿时间:2011-01-17
修稿时间:2011-02-24

IIHT: New improved iterative hard thresholding algorithm for compressive sensing
ZHANG Zong-nian,LI Jin-hui,HUANG Ren-tai.IIHT: New improved iterative hard thresholding algorithm for compressive sensing[J].journal of Computer Applications,2011,31(8):2123-2125.
Authors:ZHANG Zong-nian  LI Jin-hui  HUANG Ren-tai
Affiliation:1. School of Electronic Engineering, Dongguan University of Technology, Dongguan Guangdong 523808, China
2. Network Center, Dongguan University of Technology, Dongguan Guangdong 523808, China
3. School of Computer Science, Dongguan University of Technology, Dongguan Guangdong 523808, China
Abstract:To overcome the shortcomings of the overdependence on the measurement matrix, the high computation complexity, the long computation time of the Iterative Hard Thresholding (IHT) algorithm, a new improved iterative hard thresholding (IIHT) algorithm was proposed by studying the theory of signal reconstruction for compressive sensing. It improved the cost function and the selection method of step size for the IHT algorithm. The simulation results show that the proposed algorithm increases the probability of recovery and the speed of convergence and reduces the computational complexity and time.
Keywords:iteration                                                                                                                        hard thresholding                                                                                                                        compressive sensing
本文献已被 CNKI 万方数据 等数据库收录!
点击此处可从《计算机应用》浏览原始摘要信息
点击此处可从《计算机应用》下载全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司    京ICP备09084417号-23

京公网安备 11010802026262号