Determination of relative position and orientation of nanosatellites by video image analysis

The article discusses the option of using video images to determine the orientation and the position of the second nanosatellite relative to the first one by a series of snapshots made by the first nanosatellite. Optical cameras installed on the first nanosatellite are used as measuring devices. A new approach has been developed to solve the task set. The approach includes an image processing algorithm, a genetic algorithm of differential evolution and a model of motion relative to the center of mass. The approach takes into account the requirements imposed on the nanosatellite hardware that sets limitations on the computational resource of the airborne digital computer and on the quality of images obtained from the optical camera. The developed approach is more economic than similar algorithms. The approach was verified by the benchmark problem of separating two nanosatellites in a short time interval. The dependence of the coordinates of the center of mass of the second satellite on the current orientation and pixel orientation target coordinates in the shot is revealed. The paper includes graphs of the error of determining orientation and position of the center of mass of the second satellite relative to the first one. The approach demonstrates acceptable accuracy in the range of up to 10 seconds for the problem under consideration. The accuracy decreases dramatically as the distance between the two nanosatellites increases.