Monocular Vision-Based Pose Measurement Algorithm for Robotic Scraping System of Residual Propellant

Abstract

Robots are being proposed as human substitutes for scraping residual solid rocket propellant from double stirring blades after the mixing. For their navigation, a pose measurement algorithm based on a monocular camera mounted on them is proposed in this study. First, the pose parameters for the blade location are redefined according to the movement and geometry characteristics. After determining one feature point and the cylindrical geometric features of one blade, the pose parameters can be calculated by using the geometric model of the features in the camera perspective projection. Experimental results show that the measurement errors of this system are below 1.5 mm and 1.7 ° for position and attitude angle, respectively, meeting its application requirements. The proposed method, with simple mathematics, can be used to determine the pose of blades in harsh environments, which is suitable for robot-related engineering applications, and also provide references for similar objects.