Architecture Design of Typical Target Detection and Tracking System in Battlefield Environment

Aiming at the characteristics of small targets, few samples, many occlusions, and strong camouflage in the battlefield environment, a target detection and tracking system based on multi-source sensor data is designed. The system is mainly composed of three parts: the unmanned aerial vehicle (UAV) platform, the ground display control platform (GDCP), and the offline algorithm training platform. The offline-trained target detection and tracking algorithm is deployed in the UAV-borne AI processing chip, which can conduct the real-time detection and tracking of typical targets in the battlefield and send the results and attitude information of UAVs to the GDCP. The system can realize real-time detection of typical battlefield targets and tracking control through the GDCP.