A video SAR moving target detection method based on shadow detection

In video SAR (Synthetic Aperture Radar) images, the movement of the target leads to defocus and displacement, making it difficult to directly identify the target itself from the images, but the occlusion of the target leaves shadows on the relative position, which could be used to detect the moving targets. On the basis of shadow detection, this paper proposes a moving target detection algorithm, which combines fast background modeling and establishment of regions of interest based on the characteristics of video SAR images. After image registration and noise reduction on the video SAR image, an improved background modeling method is applied to effectively compensate the foreground target and quickly extract a clean background. Then, by effectively accumulating the change information of the moving target's shadow between the image sequences, the region of interest is established to reduce the subsequent detection range. In a comparative experiment conducted on the video SAR data (Eubank gate) released by Sandia Lab, the algorithm has achieved good detection performance, showing that it can effectively suppress false alarms and missed alarms, and has stronger robustness when detecting multiple targets.