Development of UAV-Based Target Tracking and Recognition Systems

Unmanned aerial vehicles (UAVs) are advantageous in their high maneuverability for long-range outdoor target tracking. In this paper, we develop a UAV-based target tracking and recognition system using an intelligent gimbal system with the capabilities of accurate camera positioning, fast image processing, and multi-modality information fusion. Algorithms of consensus-based target tracking, moving background processing, and neural network-based target detection/recognition are optimized for embedded implementation. A geographic information system (GIS) is utilized to provide geo-location, environmental, and contextual information. The experimental results demonstrate the robust performance of the proposed UAV-based target tracking and recognition framework.