A new approach for landmine discrimination in SAR images

Landmine detection is an available application of Synthetic Aperture Radar (SAR). According to the double-hump structure of a landmine in SAR images, a new approach is proposed for landmine discrimination with sparse representation. First, the total variation (TV) algorithm is used to denoise. Second, a sub-dictionary is obtained within the narrow limits of the Gabor dictionary. Third, the sparse reconstruction mean square error (MSE) is calculated for classification. The experimental results indicate that the proposed approach is effective for landmine discrimination.