Study of online 3D vision coordinate measuring system for large-scale equipments

To solve the problem of online measuring large-scale equipments, the paper proposes a stereo vision coordinate measuring system based on optical bar with the function of measuring and calibration, which overcomes the weakness of the existing vision measuring systems that instability without restriction during the measuring process and lower measuring precision. The system adopts the optical bar as the imaging object and obtains the 3D coordinate of the measured object with the images of the optical feature points on the optical bar got by the two cameras which are arbitrarily played. During the measuring process, the extrinsic parameters of the stereo vision system are calculated timly with the measuring data to monitor the measuring system state. Key technologies influencing system precision include the design of the calibration pattern and calibration arithmetic of the camera intrinsic parameters and the extrinsic parameters online calibration of stereo vision system, and stereo vision measuring model, etc. We deeply study these technologies, propose the corresponding schemes, reduce their effect and improve the system precision. The experiment result shows that the system can expediently satisfy high accuracy online measuring of large-scale equipments, and the maximum measuring error is 0.11mm.