A Novel Hybrid Approach for Radar Target Classification Based on SVM and Central Moments with PCA Using RCS

摘要

Radar cross section values are features which have been frequently used in target classification. The classification performance can be increased by extracting statistical properties of these features. In this paper, central moments are obtained from Radar Cross Section (RCS) values. Next, as a novelty Principal Component Analysis (PCA) is applied to these moments. Then the features extracted in this way are classified by Support Vector Machine (SVM). In order to compare the performance of proposed approach, the results are given according to varying SNR. In order to evaluate the effect of number of eigenvectors, the results are given by changing the number of eigenvector. Finally, the execution times and error performances of the different approaches are compared.