Vision-Based Relative State And Inertia Ratio Estimation Of Unknown Targets

This paper develops an algorithm to estimate the relative pose, motion, and inertia ratio of an unknown target using stereoscopic vision. First, the positions and velocities of detected feature points on the spacecraft are estimated. Second, the angular velocity and attitude of the target are estimated by the least square method and q-method, respectively. Third, the position and velocity of the center of mass of the spacecraft are recovered. Finally, the inertia ratio is estimated based on the angular momentum conversation using the estimated angular velocity and attitude of the target. Numerical simulations are conducted to demonstrate the proposed algorithm.