电光与控制, 2011, 18 (12): 35, 网络出版: 2011-12-16   

一种基于多特征提取的雷达目标识别方法研究

A Radar Target Recognition Approach Based on MultiFeature Extraction
作者单位
空军工程大学导弹学院,陕西三原713800
摘要
直接利用雷达高分辨一维距离像或是提取一维像单一特征进行目标识别难以取得理想效果。为了更好地反映雷达目标本身的物理特性,提高雷达目标识别率,对雷达目标高分辨一维距离像的频谱幅度、能量聚集区长度、散射中心数目和中心矩特征进行提取并构成多特征向量,描述雷达目标高分辨一维距离像频域、能量等物理特性,采用主成分分析方法进行特征维数压缩,基于支持向量机的方法进行分类识别,从而提出一种基于多特征提取的雷达目标识别方法。实验结果表明,采用该多特征提取方法进行雷达目标识别有助于提高识别率和减少分类时间。
Abstract
The direct use of high resolution radar range profile or single feature extracted from a onedimensional image is difficult to achieve a desired result for target recognition.In order to reflect the radar target characteristics of the physical structure better and increase radar target recognition rate,the spectrum range,energy accumulation zone length,the number of scattering centers and central moment features of High Range Resolution Profile (HRRP) were extracted to constitute a multifeature vector.This multifeature vector could describe the energy and other physical characteristics of high range resolution profile.The principal component analysis was used for feature dimensionality reduction,and Support Vector Machine (SVM) was used for classification and recognition.Thus an approach was proposed for radar target recognition based on multifeature extraction.Experimental results show that the multifeature extraction is helpful for improving the recognition rate and reducing the time for classification.

杨磊, 王晓丹, 张玉玺. 一种基于多特征提取的雷达目标识别方法研究[J]. 电光与控制, 2011, 18(12): 35. YANG Lei, WANG Xiaodan, ZHANG Yuxi. A Radar Target Recognition Approach Based on MultiFeature Extraction[J]. Electronics Optics & Control, 2011, 18(12): 35.

本文已被 1 篇论文引用
被引统计数据来源于中国光学期刊网
引用该论文: TXT   |   EndNote

相关论文

加载中...

关于本站 Cookie 的使用提示

中国光学期刊网使用基于 cookie 的技术来更好地为您提供各项服务,点击此处了解我们的隐私策略。 如您需继续使用本网站,请您授权我们使用本地 cookie 来保存部分信息。
全站搜索
您最值得信赖的光电行业旗舰网络服务平台!