Support Vector Machine based micro-Doppler signature classification of ground targets

In this paper, design of a micro-Doppler signature classifier for NR-V3 Ground Surveillance Radar* is discussed. The classifier distinguishes between pedestrians, vehicles and no target (noise) classes. Feature vector inputs for the classifier are extracted by preprocessing the FFT spectrum of radar backscattered signal. Support Vector Machine (SVM) with Radial Basis Function (RBF) and Polynomial kernels is used for classification of feature vectors. The classifiers are trained and tested using data collected with NR-V3 radar. This technique achieves a classification accuracy of over 94%.