Slow-Moving Target Detection of SAR images using High-IOU Sub-Aperture Decomposition

Inhalt:
To tackle the challenge for the detection of slow-moving targets in single channel SAR images, a new method is proposed to improve the detection performance and to estimates the radial velocity by using the High Intersection-over-Union Sub-aperture Decomposition (HISAD) method. Firstly, the Along Track Interferometry SAR (SAR-ATI) system is given by modelling the working modes of binary channels. Then the concept of equivalent interference baseline is proposed, and the linear relationship between synthetic aperture time and frequency of radar signal is used to align the sub-aperture images generated by the HISAD method. The feasibility of this method was validated by comparing the equivalent ATI results of the HISAD method with the real results of the two-channel. Finally, the Magnitude and Phase Constant False Alarm Rate (M&P-CFAR) detector was constructed, and its threshold expression was derived to successfully separate the slow-moving target from the clutter background. The results show that the new method effectively detect slow-moving targets from single-channel SAR images which covered with clutter, and reasonably estimate their radial velocity.