摘要
雷达目标识别的预处理工作是高分辨距离像领域中的重要组成部分,也是提高识别率的重点和难点。给出了一种结合信号统计特性的信号预处理方法,通过对高分辨距离像(highresolutionrangeprofiles,HRRP)的非线性变换作为特征,有效地拉大了异类目标信号之间的欧氏距离,从而提高了分类的识别率。基于ISAR雷达实测飞机数据的实验结果证明了该方法的有效性。
Pre-processing of radar target recognition with high-resolution range profiles(HRRPs) is a significant but difficult point in improving recognition performance. A new signal pre-processing method based on statistic features is proposed, which successfully increases the Euclidean distance among feature vectors of separate targets by a nonlinear transform of HRRPs, thus enhancing the recognition performance. Experimental results based on the ISAR radar airplane data demonstrate the improved performance with the proposed method.
出处
《系统工程与电子技术》
EI
CSCD
北大核心
2005年第10期1732-1733,1751,共3页
Systems Engineering and Electronics
基金
"十五"国防预研项目基金(413070501)
国家自然科学基金(60371044)
国家留学回国人员科研基金资助课题
关键词
高分辨率距离像
统计特性提取
自动目标识别
模式分类
high resolution range profiles
statistic feature extraction
automatic target recognition
pattern classification