Point cloud uncertainty analysis for laser radar measurement system based on error ellipsoid model

Abstract

Three-dimensional laser scanning has become an increasingly popular measurement method in industrial fields as it provides a non-contact means of measuring large objects, whereas the conventional methods are contact-based. However, the data acquisition process is subject to many interference factors, which inevitably cause errors. Therefore, it is necessary to precisely evaluate the accuracy of the measurement results. In this study, an error-ellipsoid-based uncertainty model was applied to 3D laser radar measurement system (LRMS) data. First, a spatial point uncertainty distribution map was constructed according to the error ellipsoid attributes. The single-point uncertainty ellipsoid model was then extended to point–point, point–plane, and plane–plane situations, and the corresponding distance uncertainty models were derived. Finally, verification experiments were performed by using an LRMS to measure the height of a cubic object, and the measurement accuracies were evaluated. The results show that the plane–plane distance uncertainties determined based on the ellipsoid model are comparable to those obtained by actual distance measurements. Thus, this model offers solid theoretical support to enable further LRMS measurement accuracy improvement.