Calibration and uncertainty estimation of non-contact coordinate measurement systems based on Kriging models

Non-contact three dimensional (3D) coordinate measurement systems (CMSs) using optical scanning techniques have the advantage of fast acquiring large numbers of points. However, these systems are known to be less accurate in comparison to the contact-based counterparts. To improve the measurement accuracy of non-contact 3D CMSs, a novel Kriging models-based calibration and uncertainly estimation method is proposed. The spatial correlation of measurement uncertainty is investigated and the calibration values for unmeasured points based on Kriging models are estimated. A procedure of best model selection is presented, and the influence of the calibration model parameters is analyzed. The proposed calibration method is validated through an ATOS II triple scan system. The results show that a significant accurate improvement for the non-contact 3D measurement system is achieved.