Vision-based simultaneous measurement of manipulator configuration and target pose for an intelligent cable-driven robot

Cable-driven robots have great application potentials in confined spaces due to their light and slim bodies. However, their kinematic accuracy is generally much lower than that of the traditional articulated manipulators, because it is difficult to install sensors on its joints to directly measure the joint position. To solve the above problems, this paper proposed a vision-based method for cable-driven robots to simultaneously measure the manipulator configuration and the target pose. The configuration is reconstructed through the detection of Apriltag 2D code and the kinematics model of cable-driven robots. The target detection method in this paper consists of the template matching method and the natural features detection. Experiments are carried out to verify the method proposed in this article, including kinematic parameters calibration, arm-shape measurement, target pose measurement and docking test. The results show that the average position error of the robot end-effector decreases from 9.12 mm to 1.86 mm after calibration. The links’ position errors measured by the method in this paper are less than 10 mm. The position and the angle errors are respectively 2 mm and 1° at the target pose measurement, and the success rate of docking experiment is more than 98%. The experimental results show that the simultaneous measurement method of manipulator configuration and target pose proposed in this paper has a great effect on improving the motion accuracy and the working ability of the intelligent cable-driven robot.