Automatic Calibration Algorithm of Robot TCP Based on Binocular Vision

Abstract

The Tool Center Point (TCP) coordinate parameter of the robot end tool is the benchmark of the robot operation. In this paper, a robot TCP automatic calibration algorithm based on binocular vision measurement is proposed. A target which can be recognized by the binocular vision sensor is attached to the robot TCP. The pose transformation between the vision sensor and the robot base is calculated by taking the binocular vision three-dimensional space measurement as the constraint and combining with the multiple translational motions of the robot end tool. After several free rotations of the end tool of the robot, TCP takes the measurement vector of the corresponding binocular vision sensor as the stroke to carry out the hypothetical parallel movement. So that the translated TCP is located at the same point in the space. To calculate the data of the corresponding flange coordinate system after the translation of TCP, and finally infers the position parameters of the TCP of the robot.