Applications of kernel methods to polarization radar target recognition using polarizationscattering matrix

In this paper, we propose novel methods of polarization radar target recognition based on kernel methods using polarization scattering matrix (PSM), in which feature extraction from PSM is avoided. Firstly two kinds of kernel function based on PSM are defined, then, they are employed to the kernel principal component analysis (KPCA) respectively. Finally the nearest neighbor (INN) classifier and the support vector machine (SVM) classifier are used for classifying targets. Experimental results based on simulated data show that the proposed methods achieve good recognition performance. Several forms of kernel function are used to prove the generalization of the proposed kernel methods. In order to validate its effectiveness, radar target recognitions of high resolution radar profiles (HRRP) are done, and the results show that the proposed methods are feasible.