Moving Target Shadow Detection and Velocity Estimation for Video SAR Based on GoDec

Video SAR has the characteristics of all-day, all-weather and high resolution, and can capture moving targets and their shadow information. For the video SAR image sequence, the Doppler shift and modulation caused by the target motion makes the moving target shift and defocus in the image, and forms shadows in the real position. These shadows can reflect the real position and state information of the target. In order to effectively realize the shadow detection of moving targets and the speed estimation, this paper introduces a GoDec-based video SAR moving target shadow detection and speed estimation algorithm for video SAR image sequences. In this method, the video SAR image sequence is used as the input, and the image is divided into the background image (stationary target) and the foreground image (moving target shadow) by the low rank and sparse feature. Then, the CFAR algorithm, the erosion dilation algorithm and the DBSCAN algorithm are used to detect and cluster the shadows. Finally, the moving target is tracked by the intersection-over-union (IOU) tracker, and its trajectory information is used to estimate the speed of the moving target. The method is verified by real w-band dataset.