Detection and Location System of Dynamic Flying Small Target Based on Vision and Radar Sensor Fusion

Giving the mobile platform the ability to detect and locate targets in real time is one of the keys to make the unmanned system autonomous. This paper presents a high dynamic small target detection, tracking and positioning system, which is composed of a visual camera and a millimeter wave radar. The system uses the target detection method based on deep learning, the target tracking method based on correlation filtering, and the target positioning method based on millimeter wave radar ranging. The system includes two modules, detection and tracking module and ranging and positioning module. The detection and tracking module performs initial detection and continuous tracking of targets based on optical images, while the ranging and positioning module realizes real-time ranging and positioning of small targets, forming a real-time high-precision dynamic flying small target detection and positioning system. Compared with UAV patrol based on ground station, this method realizes autonomous target perception in airborne/vehicle terminal. The operation speed has reached more than 10fps, and the detection rate has reached more than 95%.