Model-Based Decomposition Feature Learning With Adversarial Prior

Chen Yang, Yue Hu, Biao Hou, Bo Ren, Jocelyn Chanussot, Licheng Jiao

Abstract: Model-based target decomposition (TD) method has been widely applied due to its clear physical scattering significance. However, after establishing decomposition basis, the process of solving the scattering components and parameters is usually underdetermined, which will lead to issues such as component negative power and overestimation. For this problem, this letter examines the TD task from the perspective of deep learning and proposes an adversarial decomposition feature learning (ADFL) model. This model could learn decomposition features suitable for current terrain characteristics according to input data. At the same time, the model-based adversarial feature prior is embedded in ADFL to maintain the physical scattering meanings. On real polarimetric synthetic aperture radar (PolSAR) datasets, the learned features of the proposed model are well correlated with real terrain scattering characteristics. Furthermore, it avoids negative decomposition features and makes more accurate fitting of scattering components, effectively alleviating the above problems.