A Novel Global Calibration Approach for Multi-Viewpoint 3D Vision Measuring System

To unify the local 3D point clouds for the measured object with big size or complex structure to a global coordinate system, we usually need to adopt global calibration approach to calibrate the transformation relationship between the global binocular vision sensor (GBVS) and measuring binocular sensor (MBS) in different viewpoints in the 3D Vision Measuring System (VMS). The field of view (FOV) of the two global cameras in GBVS determines the measurement range of the 3D VMS in the global calibration process. In the traditional 3D VMS of single-station multi-viewpoint, the measurement range is limited because of the two global cameras in the GBVS are needed to have a public FOV. Then, the global calibration process is also limited. And the global calibration process is complex and easy to cause error accumulation when a laser tracker is used to replace the GBVS in the 3D VMS. In order to solve this problem, we proposed a novel global calibration approach based on multi-directional target (MDT) for a novel designed 3D VMS. It can achieve the function of global calibration for the designed 3D VMS once the MDT appears in the FOV of the global camera. Finally, we adopt two global cameras which don’t need to have a public FOV to form GBVS. It can further expand the measurement range of the 3D VMS. And the global calibration efficiency can also be improved at the same time. The feasibility and effectiveness of the proposed global calibration algorithm are verified through experiments. The experimental results show that the accuracy of the proposed global calibration algorithm can reach up to 0.08 mm after optimizing. It has broad application prospects and practical theoretical research value.