Gospodnetic, Petra; Fraunhofer ITWM; Germany
Gospodnetic, P.; Fraunhofer Institute for Industrial Mathematics ITWM; Germany
Spies, M.; Fraunhofer Institute for Nondestructive Testing IZFP; Germany
Rauhut, M.; Fraunhofer Institute for Industrial Mathematics ITWM; Germany
Session: NDT Reliablity 2
Time: 11:30 - 11:50
Image based reconstruction of surfaces is a noninvasive method which allows for further analysis and modeling of the surfaces properties and behavior under different light conditions without actual need of exposing the surface to the conditions in question. Such method helps in situations where we want to predict or analyze an impact of different deformations and light conditions on the surface properties, without having to expose or alter the surface.
Microgeometry can be reconstructed by creation of per-pixel vector fields based on BRDF (Bidirectional Reflectance Distribution Function) readings from a sequence containing multiple known image acquisition conditions (illumination and point of view). Distribution of vector orientation is further used for a simulation model which allows estimation of surface response on a 3D model of the object to be inspected.
In order to quantitatively evaluate the quality of the simulation, Fraunhofer ITWM and Fraunhofer IZFP use an â versus a-analysis to determine the Probability of Detection (POD) of defects on the simulated surface. The POD designates the probability to detect a defect which is present in the component, and will here be determined as a function of defect size a. We discuss the different approaches pursued by ITWM and IZFP for determining the POD and it’s usability as a tool for determining the quality of the image processing simulation.