Target Aspect Angle Estimation for Synthetic Aperture Radar Automatic Target Recognition Using Sparse Representation

Aspect angle estimation of the targets is of great help to computational reduction for template-based synthetic aperture radar (SAR) automation target recognition (ATR) algorithms. An effective aspect angle estimation algorithm of targets in SAR images based on sparse representation is proposed in this paper. The spare vector is firstly obtained under the dictionary which is constructed by all the training samples. And then, taking the characteristic that the SAR image sample is sensitive to the target aspect angles into consideration, the reconstruction error is calculated by each training sample according to the nonzero entry of the sparse vector. The aspect angle of the sample with the smallest reconstruction error is regarded as the final output. And the proposed algorithm does not suffer the 180 degree ambiguity. Experiments carried out on the moving and stationary target acquisition and recognition (MSTAR) datasets validate the effectiveness of the proposed algorithm.