Tracking Control of a Moving Object for Robokers with Stereo Visual Feedback

Abstract

This paper presents the visual feedback control of a moving object for robot manipulators. A stereo vision camera detects the pose of the moving object and the estimated pose is used to form the tracking error such that a visual servoing loop is constructed as the eye-in-head system structure of the humanoid robot arms called 'ROBOKER'. Kinematics of the roboker is analyzed and the Cartesian controller is constructed based on kinematics solutions. Experiments arc conducted to show the feasibility of the eye-in-head system implementation.