A Novel Method for Multi-Targets ISAR Imaging Based on Particle Swarm Optimization and Modified CLEAN Technique

This is the authors' abstract. We don't add key points for this paper.

In multi-targets inverse synthetic aperture radar (ISAR) imaging, range profiles of different target are coupled together, resulting in the failure of traditional mono-target imaging method. A novel multi-targets ISAR imaging method based on particle swarm optimization (PSO) and modified CLEAN technique is proposed in this paper. First, multi-targets are modeled as several separated group-targets in which translational motion of each target is analogous. And then, translational motion of each group-target is modeled as a polynomial, and the polynomial coefficient vector is estimated via the PSO-based iteration. Furthermore, a well-focused image of the group-target can be obtained and extracted via the proposed modified CLEAN technique. Meanwhile, each target can be segmented and extracted based on clustering number estimation and K-means clustering algorithm. Finally, better focused image of each target would be obtained through further traditional mono-target imaging processing. Experimental results verify the validity of the proposed method.