Grasping Control Method of Manipulator Based on Binocular Vision Combining Target Detection and Trajectory Planning

In order to make the manipulator more accurately identify and grasp the target object, this paper proposes a manipulator control method based on binocular vision. First, the rotation relation between lever arms of manipulator is modeled by the D-H method and the model is simplified to improve the efficiency of the rotation operation. Then, apply the Canny edge detection algorithm based on binocular vision to identify and locate the target, and calculate the position of the target in the visual coordinate system. Next, the position of the target in the manipulator coordinate system is obtained according to the coordinate conversion. Finally, plan the movement trajectory of manipulator according to the inverse kinematics solution and motion trajectory planning algorithm, and drive the manipulator to capture the target. The experimental results show that the manipulator can accurately identify and capture the target with proposed control method.