Research On Multi-band Imaging Target Tracking And Positioning System Based On UAV Platform

With the rapid development of computer vision and technology of Unmanned Aerial Vehicle(UAV),the computer vision systems of UAV platform have been widely used in civil and military fields.The target tracking and positioning system can help UAV analyze itself state through the acquired image,and control its motion.In this paper,a target tracking and positioning system of UAV platform in multi-band images is designed,the independent embedded platform is installed on the UAV to process images and data online.The target tracking module and ground target localization module are important components of the system.In the target tracking module,the existing mainstream target tracking algorithms are studied widely,a steady object tracking method based on online sample mining is proposed to improve the accuracy and robustness of the tracker,which can overcome the problems of background interference,rotation change,occlusion and etc.Experiments show that this method performs better than other tracking algorithms in public datasets.In the ground target tracking module,an image-based target location method is designed,and we analyze the factors of interference location accuracy.Optical axis calibration is used to eliminate the installation error of the sensor,the random error caused by the sensor is reduced by Kalman filter,through the experimental verification,it is found that the feasibility of the positioning method.In order to improve the stability of the UAV system,a target trajectory prediction method based on attitude angle calculation of the camera is proposed to solve the rapid motion of the target caused by the dithering of UAV,the target relocation method based on detection is adopted to deal with the full occlusion of the target,which achieves long-term tracking of the target.In the end,the designed multispectral target tracking and positioning system is tested in different scenarios.The results of experiments show that the whole system can meet the requirements of real-time processing,can accurately provide the pixel position of the moving target in image and relative to the actual position of UAV.At same time,it realizes the intelligence of UAV,and has strong practicability and application.