SAR automatic target recognition based on Local Phase Quantization plus Biomimetic Pattern Recognition

Abstract

Due to the needs of the modern battlefield awareness and military reconnaissance, Synthetic Aperture Radar (SAR) automatic target recognition (ATR) has been receiving more and more attention. In this paper, a SAR ATR algorithm based on Local Phase Quantization (LPQ) plus Biomimetic Pattern Recognition (BPR) has been proposed. There are three main steps in the proposed algorithm: firstly, a simple preprocessing procedure based on centroid location is applied to the original SAR image to extract the target's region of interest. Secondly, short-term Fourier transform (STFT) computed over the subwindows at every pixel of the preprocessed image is performed and the information of the Fourier phase spectrum is quantized as the features to overcome the target azimuth angle variation. Finally, high dimensional geometry space covering method based on BPR theory is adopted and the hyper-sausage neuron is employed to construct the hyperlinks for the purpose of recognition. Experiments on the MSTAR database show that this proposed algorithm can achieve relatively high recognition rate and high rejection rate.

SAR object classification using the DAE with a modified triplet restriction

“…As the crucial task of automatic target recognition with synthetic aperture radar images (SAR ATR), feature engineering is aimed at finding discriminative features to distinguish objects in high‐resolution SAR images [1, 2]. Although numerous papers have been published in the past decades, it is still a challenging task because of the complicated imaging mechanism, the increasing resolution of images, the complicated structures of targets and the complex land form of observing areas.…”