The Study of Video Image Processing for Small Satellites Formation Flying Experiment

摘要

This paper presents the simulation experiment on the ground, for validating the key technologies of satellites formation flying. In the experiment, machine vision method is adopted to validate autonomous controlling algorithm and relative distance measurement among satellites. In the experiment, the vision system is built for capturing the dynamic digital images. The vision system consists of a camera, a computer used for receiving video images and image processing software. The function of image processing software is as follow; Firstly, the digital image is segmented by a given threshold, for the gray value of drones is bigger than the gray value of background and other objects, by means of abundant experiments , a suitable threshold is obtained which make the targets be selected from the background images ; Secondly, the marking algorithm is designed based on the connectivity of target area, all the targets are marked by this algorithm. During the marking process, different targets are marked by different marking numbers, when all the targets have been marked, the position and count of marks which have the same marking number are counted, and the average is the target' position; Lastly, according to the marking result, the recognition software is designed in order to recognize the marked targets, and the data are intended for the formation keeping or formation reshaping of simulators.