基于时域双谱特征的雷达目标识别

雷达目标高分辨一雏距离像(High range resolution profile,HRRP)是目标散射回波在雷达径向上的投影,具有非高斯、非线性特点.文中从双谱概念出发引入时域双谱概念.分析了HRRP的时域双谱特征.为提高识别性能,定义了Fisher类可分离度函数,以类间可分离度最大化作为特征提取准则,提取一些特征双谱作为HRRP的特征向量.采用基于子空间变换的独立分量分析(Principle component analysis,PCA)和线性判别分析(Linear discriminant analysis,LDA)的特征提取方法进行了对比实验.结果表明,本文方法用于雷达目标高分辨一维距离像识别具有良好的抗喙性能和较高识别率.

Radar Target Recognition Based on Time-Domain Bispectra Feature

Radar high range resolution profile (HRRP) contains target structure information with a promising signature for radar automatic target recognition. By introducing the conception of time-domain bispectra the time-domain bispectra characteristics of HRRPs are analyzed. To improve the recognition rate, the class separability measure function is defined based on Fisher rule and some features are extracted. The features are extracted "on the fly" without a priori restrictions regarding the location or number of features. Data dimension reduction and feature extraction method, Based on principle component analysis (PCA), and linear discriminant analysis (LDA) has wide applications in the pattern recognition field. Classification tests are performed to compare performances of the method, PCA and LDA. Results demonstrate that the method has good the robust performances.