An international conference connecting people
in CAD research, education and business
Copyright (C) CAD Solutions, LLC. All rights reserved.
Proceedings of CAD'15, 2015, 113-117
Curvature-based Coarse Registration for Dimensional Metrology
Abstract. In high-precision dimensional metrology, several and multiple measurements by means of various optical and tactile probes are required to measure the whole outer shape of freeform workpieces while controlling reliability and accuracy of the results. Data obtained from one or several measuring machines equipped with one or several probing systems are characterized by different representations, different measurement uncertainties, different/limited overlapping sections and different spatial organization of points. These data are then aligned and fused into a common coordinate system using a registration technique by computing the corresponding optimal rigid transformation parameters to obtain only one set of data covering the entire measured freeform workpiece. Registration process which almost involves both coarse and fine registration is done per two data sets (scene data and model data) endowed with common overlapping sections. The coarse registration aligns roughly the two data sets with a lower resolution from a global view. The outcome alignment is thereafter optimized through the fine registration for a higher resolution. Standardized iterative methods have been established for fine registration, in quality inspection and in reverse engineering, such as Iterative Closest Points (ICP) and its variants.
Keywords. Coarse registration, Discrete curvatures, Ransac, Hough transforms, Dimensional metrology, Tomography measurements, Data fusion