Radar Automatic Target Recognition Based On High Resolution Range Profiles

Radar automatic target recognition (RATR) is a significantly developing aspect in modern radar system. This thesis mainly investigates the methods of RATR based on high resolution range profile (HRRP). At first, it makes a brief introduction to RATR. After analyzing the properties of HRRP, this thesis studies the preprocessing methods of HRRP. Then, the improvements on traditional Adaptive Gaussian Classifier (AGC) have been completed. At last, it proposes a RATR subspace recognition method. The key works in this paper are as follows: (1) Preprocessing methods of HRRP have been studied. (2) Problems of traditional AGC have been studied and relevant improvements have been given. (3) This thesis studies the feature selection method of HRRP and proposes a RATR method in target subspace based on the analysis of noise-match in HRRP space. It is difficult to avoid noise-match in RATR. In theory, target subspace has no unnecessary noise and can describe the properties of target better. Experimental results for measured ISAR data indicate that preprocessing methods of HRRP in this paper can improve recognition rate of RATR and the average classification performance of the improved AGC is better than that of the traditional AGC. The recognition results in target subspace are more veracious than in HRRP space.