Joint Multimode Cooperative Representation Classification for Vehicle Targets in SAR Imagery

In order to improve the identification performance of vehicle targets in Synthetic Aperture Radar (SAR) images, this paper proposes a Joint Multimode Cooperative Representation Classification based Classification (JMCRC) method for SAR image vehicle targets. Firstly, Two Dimensional Variational Mode Decomposition is used to decompose SAR image into multiple sub-modal components representing global information and edge information respectively, and then extracting the two-dimensional bidirectional principal component analysis ((2D)2PCA) characteristics from each sub-modal; Secondly, the Cooperative Representation Classification was extended to the JMCRC, and the original image and features of each sub-mode were combined for the Classification task. The proposed method is verified on the MSTAR dataset and a real recorded dataset, and the results show that the method proposed in this paper achieves better classification performance under the Standard Operating Condition (SOC), specific model recognition, depression angle variance and ample unbalanced experimental conditions.