taptap下载安装安卓学报

• 民用航空 • 上一篇    下一篇

基于组合特征和SVM 的HRRP 目标识别

韩萍,孙亚萍   

  1. (taptap下载安装安卓智能信号与图像处理天津市重点实验室,天津 300300)
  • 收稿日期:2015-03-04 修回日期:2015-04-07 出版日期:2016-02-20 发布日期:2016-04-13
  • 作者简介:韩萍(1966—),女,天津人,教授,博士,研究方向为图像处理与模式识别、SAR目标检测与识别等.
  • 基金资助:

    国家自然科学基金项目(61231017);中央高校基本科研业务费专项(3122014C004);taptap下载安装安卓科研基金项目(2012KYE03)

Research on HRRP target recognition based on feature combination and SVM

HAN Ping,SUN Yaping   

  1. (Intelligent Signal and Image Processing Key Lab of Tianjin,CAUC,Tianjin 300300,China)
  • Received:2015-03-04 Revised:2015-04-07 Online:2016-02-20 Published:2016-04-13

摘要:

给出了一种基于组合特征和支持向量机(support vector machine,SVM)的雷达高分辨一维距离像(high range resolution profile,HRRP)目标识别方法。该方法首先提取非相干平均距离像与中心矩特征,然后将两种特征串行组合成新的特征,最后采用SVM 分类器进行识别。根据美国运动和静止目标获取与识别(moving and stationary target acquisition and recognition,MSTAR)实测SAR HRRP 数据进行实验,结果表明在方位角未知的情况下,能够明显提高目标的识别率,是一种有效的HRRP目标识别方法。

关键词: 高分辨率一维距离像, 组合特征, SVM, 目标识别

Abstract:

HRRP(high range resolution profile)target recognition method based on feature combination and SVM classifier is presented. Firstly,the incoherent average HRRP feature in frequency domain and central moments feature are extracted respectively. Then,they are serially combined into new features. Finally,SVM is used to perform classification. Experimental results with MSTAR (moving and stationary target acquisition and recognition)SAR HRRP data sets show that the proposed algorithm can improve correct recognition rate without target azimuth,and is an effective method for HRRP target recognition.

Key words: HRRP, feature combination, SVM, target recognition

中图分类号: 

Baidu
map