论文部分内容阅读
研究了核主分量分析(KPCA,Kernel Principal Component Analysis)在高分辨雷达目标特征提取与识别中的应用。首先讨论了KPCA算法原理,然后将KPCA应用于雷达目标距离像特征提取,并采用支持向量机进行分类,提出了基于核主分量分析的高分辨雷达目标特征提取与识别方法。在核函数的选取上构造了一个组合核函数,最后用4类目标数据进行了实验,并与采用高斯核函数方法进行了比较,实验结果表明,该方法能够提高目标识别性能。
The application of kernel principal component analysis (KPCA) in the target feature extraction and recognition of high resolution radar is studied. Firstly, the principle of KPCA algorithm is discussed. Then, KPCA is applied to radar target distance image feature extraction and classified by support vector machine. A method of target feature extraction and recognition based on KPCA is proposed. In the selection of kernel function, a combinatorial kernel function is constructed. Finally, four kinds of target data are used for the experiment, and compared with Gaussian kernel function method. Experimental results show that this method can improve the performance of target recognition.