中文摘要：

随着深空在轨失效航天器数量的不断增加，针对这类空间非合作目标的在轨服务需求变得日益迫切。空间非合作目标相对位姿估计是实现交会与对接等在轨服务的前提，本文研究了一种基于序列影像立体视觉的位姿测量方法，包括图像滤波、形态学处理、边缘检测、角点跟踪、特征匹配、三维结构重建、相对位姿测量等关键技术，且通过仿真实验进行了验证。仿真结果表明，使用该方法非合作空间目标的相对姿态测量精度优于0．15°，相对位置测量精度优于8％。

英文摘要：

Since the number of on-orbit spacecraft failure in deep space is increasing, the requirement for on-orbit service o- riented to this kind of non-cooperative space target is becoming more and more urgent. Relative pose measurement is the premise on which on-orbit service like rendezvous and docking is provided. In this paper, a pose measurement method based on image se- quence of stereo vision is studied, including key technologies such as image filtering, morphological processing, edge detection, corner point tracking, feature matching, 3D structure reconstruction, and relative pose measurement. The method is verified by simulation experiments and the results show that the relative attitude measurement accuracy of non-cooperative space target is bet- ter than 0.15 degree, and the relative position measurement accuracy is better than 8%, which proves that the method is feasible and accurate.