An adaptive sampling method for STL free-form surfaces based on the quasi-Gauss curvature grid

Abstract

ABSTRACT The contact on-machine measurement can be directly used to detect and analyze the computerized numerical control machining quality of parts. To solve the problems of unreasonable distribution of sampling points and omission of sampling area in traditional sampling methods, an adaptive sampling method of the stereolithography (STL) free-form surfaces based on the quasi-Gauss curvature grid is proposed. First, on the quasi-conformal rectangular mapping region of the triangular mesh surface, a globally optimal centroidal Voronoi tessellation is generated with discrete Gauss curvature as the region density. Then, based on the quasi-Gauss curvature grid generated by subdividing the centroid of the centroidal Voronoi tessellation through a quadtree method, several types of feature points are reasonably taken as measuring points to realize the adaptive sampling. Finally, the experiments show that, for the on-machine measurement of STL free-form surfaces, the quasi-Gauss curvature grid sampling strategy can make the distribution of measuring points more reasonable and practical.