电子学报 ›› 2016, Vol. 44 ›› Issue (6): 1336-1342.DOI: 10.3969/j.issn.0372-2112.2016.06.011

• 学术论文 • 上一篇    下一篇

GTD模型参数估计的时域稀疏成分分析法

钟金荣, 文贡坚   

  1. 国防科技大学ATR重点实验室, 湖南长沙 410073
  • 收稿日期:2014-09-17 修回日期:2015-10-11 出版日期:2016-06-25
    • 作者简介:
    • 钟金荣 男,1985年7月生,广西玉林人,现为国防科技大学ATR重点实验室博士研究生,研究方向为雷达目标特性建模,图像处理,自动目标识别.E-mail:Zhong_nudt@163.com;文贡坚 男,1972年8月生,湖南宁乡人,国防科技大学ATR重点实验室教授、博士生导师,研究方向包括图像处理,自动目标识别以及摄影测量与遥感等.
    • 基金资助:
    • 教育部新世纪优秀人才支持计划 (No.NCET-11-0866)

Time Domain Sparse Component Analysis Method for GTD Model Parameter Estimation

ZHONG Jin-rong, WEN Gong-jian   

  1. ATR Key Lab, National University of Defense Technology, Changsha, Hunan 410073, China
  • Received:2014-09-17 Revised:2015-10-11 Online:2016-06-25 Published:2016-06-25
    • Supported by:
    • Program for New Century Excellent Talents in University of Ministry of Education of China (No.NCET-11-0866)

摘要:

准确高效地估计GTD模型参数对目标特性研究和目标识别有重要的意义.本文根据雷达宽带时域信号能量集中的特点,建立稀疏成分分析的时域模型,实现GTD模型参数估计.该时域模型,根据高分辨率一维像自适应地缩小散射中心分布的可能区域,缩减字典的列数;利用GTD模型的时域响应构建时域字典,并截断字典中值较小的元素使字典成为稀疏矩阵.根据模型的特点设计了一个基于正交匹配追踪的求解方法.与现有频域模型相比,时域模型的字典不但维数减少而且是一个稀疏矩阵,能极大地降低字典的数据量和模型求解计算量.通过实验验证了时域字典的性能和参数估计方法的有效性.

关键词: GTD模型, 参数估计, 稀疏成分分析, 时域稀疏

Abstract:

Estimating of geometrical theory of diffraction (GTD) model parameters accurately and efficiently is significant for automatic target recognition.In this paper, a time-domain sparse component analysis method is designed for parameter estimation of GTD model.Firstly, potential locations of scattering centers, as well as columns of the dictionary, are reduced according to the high resolution range profile.Secondly, the time domain responses of GTD model is used to construct the dictionary, and the small amplitude areas are cut off.Compared with the presented method, not only columns of the new dictionary are reduced, but also the dictionary becomes a sparse matrix.As a consequence, it needs less memory to store, which is helpful to decrease computational complexity of component analysis.In addition, a reconstruction method based on the orthogonal matching pursuit is designed for our dictionary.Finally, experimental results demonstrate the performance and efficiency of the proposed method.

Key words: GTD model, parameter estimation, sparse component analysis, time domain sparse

中图分类号: