Radar HRRP target recognition based on robust dictionary learning with small training data size

In this paper, we develop a robust dictionary learning method for radar high-resolution range profile (HRRP) target recognition, which can utilize the structural similarity between the adjacent HRRPs and overcome the uncertainty of sparse overcomplete representations. Experimental results on the measured HRRP dataset with small training data size show our method can obtain better performance than some other reconstruction algorithms based radar HRRP target recognition methods.