A Noise-Robust Radar Target Classification Method Based on Complex Probabilistic Principal Component Analysis

We develop a noise-robust radar target classification method to discriminate the moving vehicle and walking human. The traditional real-valued Probabilistic Principal Component Analysis (PPCA) model is extended to the complex-value domain for modeling the low resolution radar echoes from the ground moving targets. The denoising preprocessing is accomplished by signal reconstruction with the proposed Complex Probabilistic Principal Component Analysis (CPPCA) model, where we utilize the Bayesian Inference Criterion (BIC) to adaptively select the principal components. After denoising, a 3-dimensional timefrequency feature vector is extracted from the denoised micro-Doppler signatures of the two kinds of ground targets, and the classification is performed via Support Vector Machine (SVM) classifier. In the experiments based on the measured data, the proposed classification scheme shows the good classification and denoising performance under the relatively low SNR condition. In the real application, the advantage in SNR can effectively extend the classification distance between the target and radar.