Analyzing the Impact of Object Distances, Surface Textures and lnterferences on the Image Quality of Low-Cost RGB-D Consumer Cameras for the Use in lndustrial Applications

Beitrag bei einer Tagung
(Konferenzbeitrag)


Details zur Publikation

Autorinnen und Autoren: Schäffer E, Beck A, Eberle J, Metzner M, Blank A, Seßner J, Franke J
Herausgeber: Schuh, G.; Schmitt, R.
Verlag: Apprimus Verlag
Verlagsort: Aachen
Jahr der Veröffentlichung: 2017
Tagungsband: 7. WGP-Jahreskongress
Seitenbereich: 215-221
ISBN: 978-3-86359-555-5
Sprache: Englisch


Abstract


Low-cost RGB-D cameras are currently used in


applications in which their functionality and low price is


preferred over accuracy. As there are many approaches for


software optimization, the focus is primarily on improving the


measurement setup to increase depth image quality of popular


RGB-D sensors: we analyze the potentials of the Microsoft


Kinect v1, Kinect v2 and Intel RealSense R200 as they differ in


weight, power consumption, resolution and technology to acquire


3D-information resulting in different strengths and potentials.


Initially, we briefly explain our measurement setup, the


adjustable inputs and resulting outputs such as standard


deviation of depth values, precision and error rate. Afterwards


the results for each indicator, depending on camera sensors,


surfaces, and distances between object and camera will be


displayed. Based on our results, it is possible to quickly derive


optimal scene surroundings to improve the depth image of a


given 3D-camera before any programming effort is needed. A


more accurate depth image improves subsequent image


processing such as mapping, object or gesture tracking. This


paper states the context in which a given camera or surface


performs best depth quality, and how to link the results to


industrial environments.


FAU-Autorinnen und Autoren / FAU-Herausgeberinnen und Herausgeber

Blank, Andreas
Lehrstuhl für Fertigungsautomatisierung und Produktionssystematik
Franke, Jörg Prof. Dr.-Ing.
Lehrstuhl für Fertigungsautomatisierung und Produktionssystematik
Metzner, Maximilian
Lehrstuhl für Fertigungsautomatisierung und Produktionssystematik
Schäffer, Eike
Lehrstuhl für Fertigungsautomatisierung und Produktionssystematik
Seßner, Julian
Lehrstuhl für Fertigungsautomatisierung und Produktionssystematik


Zitierweisen

APA:
Schäffer, E., Beck, A., Eberle, J., Metzner, M., Blank, A., Seßner, J., & Franke, J. (2017). Analyzing the Impact of Object Distances, Surface Textures and lnterferences on the Image Quality of Low-Cost RGB-D Consumer Cameras for the Use in lndustrial Applications. In Schuh, G.; Schmitt, R. (Eds.), 7. WGP-Jahreskongress (pp. 215-221). Aachen, DE: Aachen: Apprimus Verlag.

MLA:
Schäffer, Eike, et al. "Analyzing the Impact of Object Distances, Surface Textures and lnterferences on the Image Quality of Low-Cost RGB-D Consumer Cameras for the Use in lndustrial Applications." Proceedings of the 7. WGP-Jahreskongress, Aachen Ed. Schuh, G.; Schmitt, R., Aachen: Apprimus Verlag, 2017. 215-221.

BibTeX: 

Zuletzt aktualisiert 2018-10-08 um 21:56