Acoustic Imaging of Underwater Moving Targets Based on Inverse Synthetic Aperture Technique

The morphological imaging and motion estimation of underwater dynamic targets are highly valuable in numerous application scenarios. However, many applications face challenges in deploying and maintaining expensive equipment, making it difficult to implement existing array and sonar-based solutions. In this paper, we propose a low-cost underwater dynamic target imaging system based on a pair of transducers(single-input single-output, SISO). We first extract target echo through Moving Target Indication(MTI) filter. The compensation for the radial distance of moving targets is executed by estimating their motion state. A joint time-frequency analysis(JTFA) method is used for instantaneous Range-Doppler(R-D) imaging. Additionally, a data augmentation method is designed to enable precise classification and identification of underwater moving targets with fewer samples. We conducted extensive simulations and practical experiments in an anechoic pool, successfully achieving acoustic imaging of moving targets with various shapes underwater. The average recognition accuracy for underwater dynamic objects with various shapes is achieved 94.28% in simulation and 79.92% in field experiment respectively.