Manipulator hand-eye calibration method based on active binocular vision

Abstract

The invention discloses a manipulator hand-eye calibration method based on active binocular vision. The method comprises the following steps that S1, a binocular vision sensor is calibrated by using a 2D plane calibration target, and a binocular vision sensor coordinate system is established; S2, the tail end of a manipulator is moved to any position, the coordinates of the tail end of the manipulator under a manipulator coordinate system are recorded, coded images are projected to the tail end of the manipulator by a structured light generator, the binocular vision sensor acquires the images, and the three-dimensional coordinates of the tail end of the manipulator under the binocular vision sensor coordinate system are calculated by utilizing coded pattern information and polar geometric constraints; and S3, the step S2 is repeated and multiple sets of coordinate data are obtained, and the relation between the manipulator coordinate system and the binocular vision sensor coordinate system is calculated, i.e. the manipulator hand-eye relation is calculated.

Description

technical field [0001] The invention relates to a hand-eye calibration method of a manipulator, in particular to a hand-eye calibration method of a manipulator based on active binocular vision. Background technique [0002] With the development of vision technology and the improvement of robot precision requirements in industry, more and more industrial robots use vision sensors as their operation and navigation tools. In order to provide navigation guidance for the manipulator, the visual sensor must first perform hand-eye calibration to obtain the positional relationship between the visual sensor and the manipulator, and then collect the coordinates of the target object in the binocular vision sensor coordinate system according to the vision sensor, and convert it into the coordinate system of the manipulator. The coordinates below provide guidance for the manipulator operation. According to the positional relationship between the visual sensor and the manipulator, the vi...

AI Technical Summary

Problems solved by technology

The Eye-in-Hand type is that the visual sensor is installed on the manipulator. The hand-eye calibration algorithm of this installation method is more complicated, but the camera and the manipulator control end can form a closed-loop control system. With the movement of the manipulator, the influence of the camera calibration error can be reduced. Therefore, this method is not very strict on the camera calibration accuracy.

The Eye-to-Hand type is that the vision sensor is separated from the manipulator. The vision sensor is installed somewhere outside the manipulator and does not change its position with the movement of the manipulator. The hand-eye calibration algorithm of this installation method is relatively simple, but the accuracy is limited by the camera calibration. It is manifested in the accuracy of the target point coordinates calculated by the visual sensor, so the accuracy requirements for camera calibration are very strict

The existing Eye-to-Hand manipulator hand-eye calibration method, such as the patent "Indoor Robot Vision Hand-Eye Relationship Calibration Method Based on 3D Image Sensor" (Application No.: 201410166077.6), adds a marker point at the end of the manipulator and uses a visual sensor to collect the marker The method of pointing three-dimensional coordinates to obtain the coordinate data used to calculate the hand-eye relationship matrix is ​​simple, but in some cases it is not suitable to add marker points at the end of the manipulator, such as the case where the end of the manipulator is sharp and small

Images