Abstract:

According to the practical background and technical requirements for autonomous rendezvous and docking (ARVD), a high accuracy target positioning method is proposed by combining Gaussian surface fitting with bilinear interpolation in view of the performance indices such as real-time and high accuracy as well as strong anti-disturbance in the process of ARVD. The analysis of target gray-level distribution is employed to carry out the bilinear interpolation and the Gaussian surface fitting in which both real target pixels and interpolated sub-pixel points can be used so that the positioning accuracy can achieve sub-pixel level for the estimation of target center position. A series of experimental results with both simulation images and real images demonstrate that the proposed method has advantages of both excellent positioning accuracy and real-time performance in that the overall positioning error can reach approximately 1/20 pixels. Therefore, the proposed method may be applied in the ARVD, space teleoperation and other relative motion fields to play a important role in engineering practices.

Key words: Cooperative target recognition, Target positioning, Gaussian surface fitting, Autonomous rendezvous and docking (ARVD)