Local histogram equalization and homomorphic filtering to improve wide-angle stereo disparity computation

Metzner S, Ramesh R, Hausotte T (2018)

Publication Type: Conference contribution, Conference Contribution

Publication year: 2018

Publisher: KIT Scientific Publishing

City/Town: Karlsruhe

Pages Range: 265 - 276

Event location: Karlsruhe DE

ISBN: 978-3-7315-0833-5

DOI: 10.5445/ksp/1000085290

Abstract

Dimensional measurements of reflective surfaces are often challenging for optical measurement systems. The use of structured light causes overexposure of the reflective regions, resulting in an incomplete measurement. Stereo vision is an effective alternate to optical measurements. Image rectification and disparity map computation steps are used to generate a 3D point cloud which can be used for dimensional measurements. When the stereo vision setup requires cameras at wide angles to the surface normal, the image pairs are affected by varying illumination, once again resulting in bad disparity. In this paper different image processing possibilities are examined at the example of a sheet-bulk metal formed workpiece to obtain an improved disparity map of the overexposed areas of the data set. A newly developed local histogram equalization method improves the disparity map for stereo vision calculations with wide-angle camera setups.

Authors with CRIS profile

Sebastian Metzner Lehrstuhl für Fertigungsmesstechnik Rahul Ramesh Lehrstuhl für Fertigungsmesstechnik Tino Hausotte Lehrstuhl für Fertigungsmesstechnik

Related research project(s)

Strategies for function-oriented optical inspection of formed precision components (A06) (TR 73 - A06) TRR 73: Manufacturing of complex functional components with variants by using a new sheet metal forming process - Sheet-Bulk Metal Forming Jan. 1, 2009 - Dec. 31, 2020

How to cite

APA:

Metzner, S., Ramesh, R., & Hausotte, T. (2018). Local histogram equalization and homomorphic filtering to improve wide-angle stereo disparity computation. In Längle, Thomas [Hrsg.] and Puente León, Fernando [Hrsg.] and Heizmann, Michael [Hrsg.] (Eds.), Proceedings of the Forum Bildverarbeitung 2018 (pp. 265 - 276). Karlsruhe, DE: Karlsruhe: KIT Scientific Publishing.

MLA:

Metzner, Sebastian, Rahul Ramesh, and Tino Hausotte. "Local histogram equalization and homomorphic filtering to improve wide-angle stereo disparity computation." Proceedings of the Forum Bildverarbeitung 2018, Karlsruhe Ed. Längle, Thomas [Hrsg.] and Puente León, Fernando [Hrsg.] and Heizmann, Michael [Hrsg.], Karlsruhe: KIT Scientific Publishing, 2018. 265 - 276.

BibTeX: Download