卫星位置视觉测量

Vision-based satellite position measurement

目前卫星轨道参数的测量主要依靠地面测控站的无线电测量,使得测量受限于测控站的观测弧段;另一方面随着深空探测的不断深入,也对现有测控方法提出了新的挑战。利用卫星上已有的相机通过景象匹配对卫星自身位置进行精确测量,可以有效摆脱地面测量观测弧段和测量距离的约束。本文提出了一种基于星载实时图与基准图匹配的视觉测轨方法:通过景象匹配建立实时图中特征点与基准图的对应关系,再根据成像关系解算卫星的三维位置。利用卫星获取的立体图像和轨道测量数据进行了验证实验,证明了本文提出的方法具有较高的精度,实现了地基无线电测轨和视觉测轨的相互验证。本文提出的视觉测轨方法可以弥补现有测轨方式的不足,减轻地面测控系统负担,在深空探测等领域具有良好的应用前景。

Satellite orbit measurement based on ground station,such as USB and VLBI,is restricted in its observation arcs.In addition,the long distance of deep space explorer mission forms a new challenge to current satellite TTC system.If the satellite can estimate its position by its own CCD camera,the time and distance constraints will be eliminated.With a view of this,a vision-based position estimation method for satellite was proposed.It registered satellite image with a reference image and then estimated the satellite＇s position according to photogrammetry.An experiment using the image,orbit data of space mission was conducted.By mutually verifying the vision method and the traditional method,it illustrates the effectiveness and accuracy.Our vision-based position estimation method is a good complement to the traditional method;it can save work of ground observation stations and have a promising application prospect in deep space exploration.