Fang Liu, Xin Liu, and Jiaqi Liu
UAV infrared detector, visibility criterion, least square method, extended Kalman Filter
An observation model of unmanned aerial vehicle(UAV) infrared detector is developed. The visibility criterion is derived via geometric relationships between the infrared detector, target and earth, the constraint of the detector’s frame angle, and the detection range of infrared detector. Because of the infrared sensor can only obtain the angular information of the target, so it is impossible for one detector to locate position of the target. The position of the target is obtained by two or more detectors. The least square method based multi-sensor target tracking method is introduced. It is important to set a filter before using the tracking data as there are observation errors caused by UAV position error, UAV attitude error, sensor pointing error and sensor image plane position error. The extend Kalman Filter (EKF) algorithm is used to smooth the tracking signals. Finally, two UAV infrared detectors observing a low-orbit target is simulated for their observation performance. The result shows that the infrared detectors can observe the target within the scope of the detection range. When the target flies off the top of the UAV, the detectors can’t observe it because of the constraint of the frame angle. The observation error is reduced substantially after the EKF is used.