Determining returning spacecraft velocity, using image features tracking

In this article, RLS method is used to process the information obtained from the tracking of features in order to measure the linear and angular displacement of the camera between image frames and subsequently to measure the speed. On the other hand, camera image registration with reference images is used to measure the INS error in position and attitude. The EKF is used to integrate the measurements of VS with the outputs of the INS. As the accuracy of the VINS is improved, it is possible to achieve a suitable approximation of the camera movement in the time intervals between the image frames. Using this approximation as an initial point leads to the convergence of the RLS to the acceptable answer. To show the effectiveness of the proposed method, the return of a spacecraft and the images of the camera mounted on it are simulated. The INS reports the speed, position and attitude erroneously. By tracking opportunistic features and determining the location of a priori mapped landmarks in image, the errors of the INS are estimated and corrected. The results are validated by comparing them with the values obtained from solving the motion equations as error-free values. The results show that in the defined scenario for the return of the spacecraft and specifications considered for the camera and inertial sensors, the accuracy of 0.02 m/s, 5m and 0.01 deg. respectively in determining the speed, position and attitude is achievable.