北京邮电大学学报

  • EI核心期刊

北京邮电大学学报 ›› 2016, Vol. 39 ›› Issue (1): 117-121.doi: 10.13190/j.jbupt.2016.01.022

• 研究报告 • 上一篇    下一篇

广义半迭代硬阈值追踪算法及在“鬼”成像中的应用

赵生妹, 唐文娟, 郑宝玉   

  1. 南京邮电大学信号处理与传输研究院, 南京 210003
  • 收稿日期:2015-03-21 出版日期:2016-02-28 发布日期:2016-01-29
  • 作者简介:赵生妹(1968-),女,教授,博士生导师,E-mail:zhaosm@njupt.edu.cn
  • 基金资助:

    国家自然科学基金项目(61271238,61475075,61271240);教育部高等学校博士学科点专项科研基金项目(20123223110003)

Generalized Semi-Iterative Hard Thresholding Pursuit Algorithm and Its Application in Ghost Imaging

ZHAO Sheng-mei, TANG Wen-juan, ZHENG Bao-yu   

  1. Institute of Signal Processing and Transmission, Nanjing University of Posts and Telecommunications, Nanjing 210003, China
  • Received:2015-03-21 Online:2016-02-28 Published:2016-01-29

摘要:

提出了一种广义半迭代硬阈值追踪重建算法,综合了广义硬阈值追踪重建算法和半迭代思想的优点,修正了目标函数寻求最优解的搜索方向,获得了多项式的加速收敛,且不需要已知信号稀疏度.数值仿真结果表明,该算法在重构概率、峰值信噪比、信噪比、匹配度等方面的性能明显提高,且在"鬼"成像中的应用性能明显优于广义硬阈值追踪算法.

关键词: 重构算法, 硬阈值追踪, 半迭代, "鬼"成像

Abstract:

A recovery algorithm, named generalized semi-iterative hard thresholding pursuit, was proposed. The algorithm modifies the searching direction of objective function with semi-iterative idea, has a polynomial acceleration convergence, and can be used for the signal without sparse information. It is shown that the proposed algorithm did improve the performance in term of reconstruction probability, peak signal-to-noise ratio, signal-to-noise ratio, and matching rate. The reconstruction quality in ghost imaging with the proposed algorithm is superior to that with the generalized hard thresholding method.

Key words: recovery algorithm, hard thresholding pursuit, semi-iterative, ghost imaging

中图分类号: