Aufgabenangepasste, kontrollierte Oberächenextraktion aus 3D-Computertomographiedaten

The industrial computer tomography (CT) is a modern instrument for quality assurance, which enables the non-destructive and complete inspection of parts. As a result, it provides 3D volume data with all geometrical and material-based structures, which constitute the basis for a variety of applications in quality control. Therefore, the CT is already used today in many stages in the product development process for quality control, but also increasingly for the dimensional measurement of components. In order to perform such measurement applications, it requires the determination of 3D surface data, which describe the material limits of the part. Current standard methods show decisive disadvantages in practice, which is the reason that the CT is nonapplicable for many types of components and tasks or additional processing steps are required to meet specific needs. In this work a novel approach for extraction of 3D surfaces from CT data is developed, which allows the determination of the limiting surface region-dependent and adapted to the measuring or testing task. The aim is that the user gets the appropriate 3D surface model directly to the relevant requirements for his metrological problem. Both pre-and post-processing steps are superfluous here. The number of input parameters are to be kept easy to use and, if possible, automated. This method of controlled, task-adapted surface extraction from volumetric CT data provides the ability to solve a variety of measurement problems for different types of components and thus fulfills current and future challenges of the industrial CT technology