Adaptive statistical model for radar HRRP recognition

Radar automatic target recognition (RATR) should have a robust recognition performance in different noisy conditions. Most of the algorithms in RATR are based on high signal-to-noise (SNR) condition, not consider the recognition performance in low SNR. In this paper, based on the probabilistic principal component analysis (PPCA) model, we develop an adaptive statistical model for radar target recognition. This algorithm makes the parameters of PPCA model altered by different noisy conditions. Experimental results for measured data show that the average recognition performance of the proposed adaptive statistical model has an obvious improvement in low SNR.