A new method of radar target recognition based on time-frequency analysis

Abstract

In this paper, focused on three targets with simple structure, such as cone, sphere and cone, cylinder and cone, dynamic RCS echo signals is predicted with the improved Greco software. The Dynamic RCS echo signals of targets with micro-movement are preprocessed by using short time Fourier transform, and high-resolution time-frequency distribution images are obtained. A new radar target recognition method based on time-frequency distribution images is proposed. Pixel scan technique is adopted to extract the feature vector, and the nearest neighbor classifier is used for classification decision. Compared with the target recognition method based on singular value decomposition, the correction probability for recognition of the proposed method is higher. Simulation results verify the correctness and the effectiveness of the proposed method.