High-Resolution ISAR imaging based on deep unfolded Bayesian network

Under complex observation conditions, high-resolution ISAR images can be attained by the compressive sensing technique. However, the conventional methods require numerous iteration steps and time-consuming parameters tuning. To address this issue, we propose an ISAR imaging method using a deep unfolded Bayesian network. Firstly, the signal model and probabilistic model are constructed. After that, the iterative algorithm is derived. On this basis, the deep unfolded network is proposed to achieve accurate reconstruction. The experimental results indicate that well-focused imaging can be obtained by the proposed method in low SNR and incomplete data conditions.