Abstract: An autonomous detection and tracking method for rendezvous and proximity of an uncooperative target in space is proposed. After the camera captures the original images, the derived high precision satellite attitude is used to determine the precise camera’s pointing, and the initial stellar objects removal is implemented in conjunction with the stellar library, and then the target is accurately identified based on the difference in the trajectories of different objects in the images. Under the condition that only the target’s line of sight angle is observable, a relative motion equation is established based on the relative orbital elements, and an extended Kalman filter (EKF) is used to construct an angles only relative navigation filter to derive the relative motion state between the target and the active satellite. A semi physical simulation platform is built to test the performances of the method in three orbital scenarios. The results show that when the pointing accuracy of the satellite attitude determination is better than 3″, the target detection accuracy of the method can reach above 96%, and the relative orbit determination accuracy of the proposed angles only relative navigation filter can reach within 1% of the measured range.

Key words: Uncooperative target detection, Autonomous rendezvous, Visual tracking, Angles only relative navigation, Relative orbital elements