Machine Vision Calibration for Industrial Robots Target Detection

Abstract

Industrial robots are increasingly used in various 3ields of industrial production. At present, the working mode of most industrial robots is 3ixed and single, and when the working environment or target position changes, industrial robots cannot accurately grasp and need to be repositioned. In order to improve the autonomy and intelligence of its work, machine vision is gradually applied to combine with robot technology to give robots the ability to perceive the environment.With the increase of industrial demand, rapid and accurate recognition and positioning of objects with arbitrary posture in complex space environment has become a new research direction.To solve this problem, this paper studies hand-eye calibration and pose recognition in 3D vision system detection based on depth camera, and realizes industrial robots to grasp objects based on pose results of vision system. In order to solve the problem of mapping between the camera coordinate system and the industrial robot coordinate system, this paper proposes a robot calibration method, which can quickly and accurately obtain the transformation matrix between the image coordinate system, the camera coordinate system and the robot coordinate system. The camera calibration process and the imaging system model are analyzed, and then image processing algorithms such as image segmentation, target classi3ication, image matching, corner detection, target localization and tracking are used. Finally, the pose data of different positions of targets are obtained through experiments, and the error analysis of the data is carried out, which provides strong data support for accurate positioning and grasping of targets.