Fast Dynamic ISAR Imaging Method Based on Low-rank Tensor Decomposition with Alternating Minimization

Abstract

Due to the ability to continuously image and monitor, inverse synthetic aperture radar (ISAR) has been widely studied in recent years. However, the large amount of ISAR data causes a low real-time imaging performance and limits its application. Hence, we propose a fast dynamic ISAR imaging method based on low-rank tensor characteristics for continuous monitoring of moving targets in this paper. By using the multi-frame redundancy characteristic of dynamic ISAR echo data, the three-dimensional (3D) low-rank tensor imaging model is established, and the imaging problem is transformed into a problem of low-rank tensor recovery. Then, the alternating minimization (AM) algorithm is modified to figure this problem out. Experiments show that the proposed fast dynamic ISAR imaging method has high imaging quality and less computation cost compared with traditional methods.