ISAR-Diff: A High-Resolution ISAR Imaging Method Based on Conditional Diffusion Models

Due to low signal-to-noise ratio (SNR) and incomplete echoes, the performance of inverse synthetic aperture radar (ISAR) imaging tends to degrade rapidly. To address this issue, this paper proposes a high-resolution ISAR imaging method based on a conditional diffusion model, dubbed as ISAR-Diff. Specifically, an ISAR imaging diffusion model is trained. Then, the complex-valued ISAR image is then inferred using the diffusion model with a guidance function, conditional on complex-valued echoes. In particular, an ISAR imaging guidance function is designed to make ISAR-Diff applicable to different scenarios and to improve the reconstruction accuracy. Experimental results based on Monte Carlo simulation and measured data demonstrate the superiority of ISAR-Diff in low SNR and incomplete echo scenarios.