Motion Parameter Estimation and Synchronous Approximation Scheme of Non-cooperative Target

With the increasing intensity of human space activities, it is urgent to eliminate large non-cooperative targets such as failed spacecraft. In this paper, a machine vision-based method for estimating the motion parameters of non-cooperative target is proposed to meet the application requirements of failed satellite capture, and is verified by a synchronous approximation experiment. Firstly, an improved method based on region growth algorithm is proposed to calculate the spin rate of non-cooperative target. Secondly, compared with traditional 3D reconstruction method, a modified scheme based on monocular vision is designed to match feature points and reconstruct rotation axis. Finally, a simulation in Pybullet and a physical experiment based on the robot arm experimental platform are performed to validate our method. The result shows that the proposed method displays stronger adaptability to light environment and higher accuracy with less computational complexity.