Multi-parameter Vision Measurement Method And Simulation Verification Of Satellite Formation Physical Simulation System

Satellite formation flying has been considered as an inevitable direction of future space application. The ground simulation is a basis of the research on its related technologies. In order to carry on ground simulation to simulate various formation flying missions, it's needed to measure the position and attitude of the satellite simulator in real time. The control system utilizes these measurements. Since the computer vision measurement is non-contact, easily implementable, and highly accurate, considering the requirements of measurement of formation flying simulation, this paper has investigated multi-objective multi-parameter vision measurement method and the Specific studies are as follows.The vision measurement system needs to identify more than one satellite simulator, and to determine the position and attitude of each satellite simulator. In order to facilitate the identification of the satellite simulators and to get the satellite position and attitude parameters, the surface of each satellite simulator is an information-encoding measurement target. As the circular structure is simple, and its projection center is invariant, a circular target structure has been designed. To determine the binary-coded-targeting design method, the binary encoding methods, the coding capacity and the positive direction are discussed. The paper introduces the code Hamming distance. According to the coding error detection capability from the code Hamming distance, a binary code scheme with fault tolerant has been proposed. It can enhance the robustness of the vision measurement system.To identify multiple encoding targets and to get the position and attitude of each satellite simulator, multi-encoding-target image segmentation problem has been discussed. Graph-based coding target clustering segmentation algorithm has been proposed. The simulation shows that this algorithm is very effective. And the algorithm is also applicable to the cases of encoding target code information missing. This paper uses the Least-squares fitting method to fit the ellipse with the feature points of the same coding target, the method making use of this elliptical target information to get its position and attitude parameters is proposed. The simulation shows that this method can determine the position and attitude parameters of the encoding target accurately.Finally, the multi-objective multi-parameter visual measurement method based on the encoding target has been verified. To solve the problems that the segmentation algorithm for the encoding target is more complex and the attitude parameters of the encoding target are multi-value, the target state estimation method of encoding target position and attitude parameters is proposed. The method simplifies the complexity of the algorithm and improves its efficiency, but also makes the attitude parameters of the encoding target to uniquely identify. The experimental results show that this measurement method can well meet the visual measurement tasks.