Optimal Kernel Discriminant Analysis for Radar Target Recognition

Kernel discriminant analysis (KDA) is an effective method for nonlinear feature extraction in radar target recognition, but it is usually difficult to solve due to the singular problem. Based on the idea of subspace projection, an optimal kernel discriminant analysis (OKDA) is given and used to extract features from a range profile. Then, kernel-based nonlinear classifiers are applied for classification. Experimental results on both simulated and measured profiles show comparatively good recognition performance of the proposed method.