Novel approach of SAR target classification utilizing aspect information

Aiming at the computational burden of template-based SAR target classification, an approach of SAR target classification utilizing aspect information is presented. First, estimate target aspect from peaks. Then consider the aspect as cueing information to reduce the search space. Finally the candidate target is declared by matching against those templates under aspect constraint. The proposed method is evaluated on MSTAR SAR target imagery, and the experimental results indicate that it achieves improved efficiency as well as comparable performance, compared with direct correlation method.