UNDERCOMPLETE DICTIONARY-BASED FEATURE EXTRACTION FOR RADAR TARGET IDENTIFICATION

Feature extraction is a challenging problem in radar target identification. In this paper we attempt to exploit the sparse property of the scattering signature with a undercomplete dictionary for target identification, and establish a feature extraction scheme based on the undercomplete dictionary. Furthermore, as an application, we present a feature vector, named as the atom dictionary feature, which is extracted from the scattering signatures over a wide-angle sector. Numerical simulation results show that the proposed atom dictionary feature can improve the performance of radar target identification due to the exploitation of the sparse property of the scattering signature.