A dense vehicle target detection method based on CSAR image
Guocheng Li, Daoxiang An · 2022
The traditional synthetic aperture radar (SAR) image can only detect a piece of target area, but can not accurately locate a single vehicle target in the target group. In order to improve the accuracy and efficiency of dense vehicle detection in SAR images, a method of segmentation detection for dense vehicle targets was proposed to make full use of the advantages of CSAR (Circular SAR) in acquiring target scattering information from all directions. The method separates the entangled adjacent targets through image segmentation, and then uses the clustering method based on azimuth angle to locate the target precisely. Experimental results show that the detection probability of the proposed method is 93.18% and 91.84% respectively for high-frequency and low-frequency CSAR image dense vehicle targets, which can locate the densely packed vehicle targets in the scene more accurately, reduce the false alarm rate, and realize the detection of dense vehicle targets. Compared with other dense vehicle target detection methods, the detection performance of this method is better.