Vision-based robot automatic calibration method, equipment and storage medium

Abstract

The invention discloses a vision-based robot automatic calibration method, equipment and a storage medium. A MARK point is fixed at the tail end of a robot, and the attitude of the MARK point is observed with a camera. The calibration method includes the steps that (1) the robot moves by nine step lengths to obtain nine point locations which are in a 3*3 manner, position coordinates of the tail end of the robot in a robot coordinate system and pixel coordinates of the MARK point in the camera coordinate system are recorded in correspondence with all the point locations, the recorded position coordinates and pixel coordinates of the nine point locations are processed according to a nine-point calibration method, and a mapping matrix X3*3 between the robot coordinate system and the camera coordinate system is obtained; (2) according to the mapping matrix X3*3, a tool coordinate system T(Xt,Yt,theta) is calculated and established with the MARK point as the center; and (3) a conversion relationship between the pixel coordinates and the position coordinates of MARK point is calculated according to the tool coordinate system T and the mapping matrix X3*3, and nine-point calibration of the camera coordinate system to the robot coordinate system is achieved. The calibration method does not require manual teaching, simplifies the calibration process and effectively improves the calibration speed and the calibration accuracy.

Description

technical field [0001] The invention relates to the technical field of machine calibration, in particular to a vision-based automatic calibration method, equipment and storage medium for a robot. Background technique [0002] In conventional 2D vision applications, take eye-to-hand as an example, refer to figure 1 As shown, the camera is fixedly installed relative to the manipulator, so the camera coordinate system is fixed relative to the workpiece coordinate system. The purpose of the nine-point calibration is to obtain the conversion matrix between the pixel coordinates and the end coordinates of the manipulator. [0003] Assuming that the image coordinates are (x, y) and the manipulator coordinates are (X, Y), the manipulator coordinates can be regarded as obtained from the image coordinates through rotation and translation, as shown in formula (1): [0004] [0005] And then there is, [0006] [0007] For different points, there are: [0008] [0009] Writ...

AI Technical Summary

Problems solved by technology

[0016] From the perspective of the calibration process, it is necessary to print a nine-point calibration paper of appropriate size according to the camera's field of view, install a calibration pin at the end of the robot, and manually teach nine robot coordinates, which is cumbersome and cumbersome to operate.

Images