Radar high resolution range profile target recognition based on T-mixture model

This paper develops a T mixture model (TMM) approach to automatic radar target recognition. The specific problem considered is the classification of radar high-resolution range profile (HRRP). We model the radar returns from each target as a T-mixture distribution. Estimation of the model parameters is achieved using a modified expectation-maximization (EM), i.e., fuzzy EM (FEM) algorithm. Experimental results demonstrate the proposed statistical model not only has better recognition performance but also is more robust to noises than other methods.