Tracking of Moving Target Based on SiamMask for Video SAR System
SiamMask outperforms KCF for tracking moving targets in video SAR imagery, and target shadows assist tracking because they remain undistorted.
Published Dec 1, 20197 citationsPaper ↗
Only vote on papers you've read. Sign in to vote.
Abstract
Video Synthetic Aperture Radar (Video SAR) technology has developed rapidly in recent years, realizing the automatic tracking of moving targets has gradually become the focus of research in the field of SAR. In this work, “SiamMask”, the newly included algorithm in CVPR2019 with the highest accuracy in the field of computer vision, is applied to the tracking of Video SAR images. Firstly, the high-quality SAR images of the imaging area are obtained, and then on the basis of these data, the initial position of the moving target is acquired. Finally, the effect of SiamMask in tracking teat data is compared with the effect of the traditional tracking algorithm - KCF. The result shows that SiamMask algorithm performs better than the traditional one in dealing with test data. In addition, since the moving target's shadow does not defocus or shift, it can be used to assist tracking. The shadow data set used also verifies the feasibility of tracking moving target's shadow, which provides ideas for subsequent research.