Abstract:

In order to improve the performance of the synthetic aperture radar (SAR) automatic target recognition system, a new method is proposed for target aspect estimation. A simple adaptive threshold processing is used to extract the strong scatters in target area. By analysis on the projection distribution in different  directions of the strong scatters, the normal front edge response is defined and used as a basis of the aspect estimation. The 90degree correction is carried out on several unreliable results in the last step. Experiments are conducted on the moving and stationary target acquisition and recognition (MSTAR) public data set, and the estimating errors are less than 10 degrees for 99% of the samples using the new method. According to the experimental results, the proposed method can achieve the optimal performance compared with the major edge fitting method, while has simpler processing flow and higher computing efficiency.