Pose Measurement of Non-cooperative Spacecraft by Sensors Fusion

The pose measurement of the non-cooperative spacecraft is one of the key technologies for the autonomous navigation systems, and the fusion of multi-sensors can improve the pose measurement accuracy. This paper proposes a pose measurement method for a non-cooperative target by fusing the information of two kinds of sensors, binocular vision and 3D LiDAR. Specifically, the line features are firstly detected from the vision to estimate the rough relative attitude of the target. Then the point features of the binocular vision are combined with point cloud of the 3D LiDAR to acquire the rough relative position of the target. Furthermore, the available pose is used as the initial value of point cloud matching to acquire a higher precision result, as well as avoid the local minimum in iterative closest point. Finally, numerical simulations are illustrated to verify the reliability and robustness of the method.