Research on Target Positioning and Tracking Method of Land-Based Flight Test Monitoring PTZ

Facing the background of the reform of the traditional flight test mode, which is characterized by “digital intelligence empowered test flight”, there is a need to develop a vision for building intelligent automation capabilities. To address the specific requirements of high-definition image video recording for the near-field road surface section of test flight subjects, a method for localizing and tracking the target of land-based PTZ test flight has been proposed. Utilizing the GPS/BeiDou dynamic differential computation model, the system calculates real-time relative orientation data for the target. This data is then used to guide the PTZ's rotation and adjust the focus of the high-definition camera, facilitating the automatic capture of the target's field of view. Subsequently, the YOLOV5s target detection algorithm, which incorporates enhancements such as global attention mechanism (GAM), dynamic head, and Wise-IoU (WIoU), performs real-time target detection and identification. Finally, in order to achieve continuous and real-time tracking of targets, a combination of median flow, Kalman filter, and IoU threshold detection techniques is employed. Extensive experimentation and comparative analysis confirm the effectiveness of this method in terms of real-time performance and tracking accuracy. Consequently, this approach satisfies the demanding requirements of practical test flight engineering applications.