基于局部积分双谱与SVM的雷达目标识别

针对高分辨距离像的特点,应用双谱与SVM研究高分辨雷达目标识别问题。提出了基于局部积分双谱与SVM的雷达目标识别方法。该方法选择具有最强鉴别能力的积分双谱构成局部积分双谱特征,基于局部积分双谱进行距离像特征提取,然后应用支持向量机对提取的特征进行分类识别。利用四种局部积分双谱的目标识别仿真实验结果表明,提出的方法具有良好的分类性能。

Recognizing Radar Targets by Local Integral Bispectrum and SVM

Based on the characteristic of HRRP,radar target recognition is studied by using bispectrum and SVM. The method for radar target recognition using local integral bispectrum and SVM is proposed. First,it extracts HRRP by choosing the integral bispectrum which has the best differentation ability to conform a local integral bispectrum,and then,classify their features with a multi-class recognition method based on SVM. Some simulations with four type local integral bispectrum have been done,and the results show that this method is good for classfication.