A satellite-borne SAR target recognition method based on supplementary feature fusion

摘要

The spatial information processing of satellite-borne synthetic aperture radar (SAR) images has very important meanings. This paper proposes an SAR automatic target recognition (ATR) method based on the fusion of complementary features. PCA features, elliptical Fourier descriptors (EFDs) and local binary pattern (LBP) are used to describe SAR images from different aspects thus they can jointly give the SAR targets more detailed representations. The three features are classified by sparse representation-based classification (SRC), respectively. And their decisions are fused based on Bayesian decision fusion for target recognition. Experiments are conducted on the moving and stationary target acquisition recognition (MSTAR) dataset to evaluate the effectiveness of the proposed method.