Joint sparse representation for multi-resolution representations of SAR images with application to target recognition

In this paper, we propose a target recognition method for synthetic aperture radar (SAR) images based on multi-resolution representations. Considering that the discriminability of SAR targets may lie on different resolutions, the original images at a fixed resolution are used to generate the multi-resolution representations of the target. Then a multi-resolution dictionary is built, which includes the multi-resolution representations of the training samples. For the test sample, its multi-resolution representations are jointly classified using the joint sparse representation (JSR) model based on the multi-resolution dictionary. The multi-resolution dictionary can not only augment the representation capability of the dictionary but also enhance the robustness of the representation. Furthermore, the JSR can exploit the inner correlations among the multi-resolution representations of the test sample. Therefore, a more precise representation of the test sample can be obtained, which will effectively improve the recognition performance. Experiments are conducted on the moving and stationary target acquisition and recognition (MSTAR) data-set to validate the effectiveness and robustness of the proposed method.