Automatic Radar Target Recognition Based On High Resolution Range Profiles

This thesis mainly investigates the methods of radar target recognition based on High Range Resolution Profile (HRRP). Firstly, it gives a survey on the radar target recognition and takes a brief introduction to the foundational of radar target recognition. Secondly, it introduce the basis concept and the characteristic research of the radar range profile, and then research the methods of feature extraction and the design of the classifier. The mainly work of this paper as follows: (1) A new signal pre-processing method based on statistic features is given in this paper, which successfully increases the Euclidean distance among different signals by nonlinear transform of High-Resolution Range Profiles, thus enhancing the identification rate of classification. (2) Radar high-resolution range profile is very sensitive to target-aspect variation. In this paper, dynamic aspect warping is presented to tackle this aspect sensitivity for high resolution range profile radar target recognition. Experimental results of the proposed method for real-life ISAR data target recognition and their comparison with the RBF and the MCC-TMM method indicate effectiveness and efficiency of the proposed method.