Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

Towards light transport matrix processing for transparent object inspection

: Meyer, J.; Längle, Thomas; Beyerer, Jürgen

Postprint urn:nbn:de:0011-n-4801662 (1.0 MByte PDF)
MD5 Fingerprint: 9b03d3d4ca16b2d2d1e7f27d4bc863b2
© IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE.
Created on: 16.1.2018

Institute of Electrical and Electronics Engineers -IEEE-; Science and Information Organization -SAI-, Bradford:
Computing Conference 2017. Proceedings : 18-20 July 2017, London, United Kingdom
Piscataway, NJ: IEEE, 2017
ISBN: 978-1-5090-5442-8
ISBN: 978-1-5090-5443-5
ISBN: 978-1-5090-5444-2
Computing Conference <2017, London>
Conference Paper, Electronic Publication
Fraunhofer IOSB ()
camera; light source; inspection; optical imaging; optical scattering; feature extraction

Objects made out of transparent materials play important roles in human's everyday life. The majority of applications that employ transparent materials require them to meet high quality standards. Particularly, transparent materials have to be free from so-called scattering defects, e.g., enclosed air bubbles. Common dark field setups can generally be used to automatically test transparent objects for such defects. However, their adaption to the concrete test object on hand often represents a time consuming task. This contribution introduces a method that combines an optical system consisting of a telecentric camera and a spatially programmable area light source with the theory of light transport matrices. Two features are presented that can be extracted out of these matrices and that allow to image scattering defects present in a transparent object without the need of adapting the system to the actual test object. A physically based rendering framework is adequately extended so that light transport matrices can be efficiently approximated for synthetic inspection scenes. By this means, the proposed approach could be successfully evaluated and it could be shown that it even outperforms a conventional visual inspection setup in some situations.