Abstract: A vision system of space robot can not observe a whole feature image of a large non-cooperative target to measure the relative position and attitude. To overcome this problem, a point structured light vision system composed by a point structured light and a single camera was established to determine the relative pose. An unknown size partial rectangular framework of the target was chosen as the measurement object. Firstly, themeasurement model of the point structured light vision system was built and the relationships were presented. Secondly, according to the intrinsic constraint of the point structured light vision system, a whole feature was reconstructed and four feature points were calculated. Thirdly, the relative transform matrix between the measurement system and the target was computed by the  four feature points. And the relative position and attitude was derived from the transform matrix. Finally, numerical simulations were studied to verify the method and the measurement errors were analysed. The results show that this method is effective.

Key words: Point structured light vision, Non-cooperative large target, Feature reconstruction, Pose measurement