Seite - 169 - in Proceedings - OAGM & ARW Joint Workshop 2016 on "Computer Vision and Robotics“
Bild der Seite - 169 -
Text der Seite - 169 -
Localization
of
an Automated Guided Vehicle (AGV) by
Stereo Based
Visual Odometry and
Artificial Landmark
Detection
Daniel Klingersberger
and Gerald Zauner
FH
OÖ Forschungs & Entwicklungs
GmbH, Stelzhamerstr. 23, 4600 Wels
gerald.zauner@fh wels.at
Abstract
Localization in a known environment is an essential topic in the field of robotics – consequently a
variety of methods (e.g. Visual Odometry or SLAM) are scientifically well established. Compared
to experimental robotics, the determination of position based on Machine Vision approaches is not
yet fully implemented in the domain of Automated Guided Vehicles (AGV). Thus, the aim of this
master’s thesis is to design, realize and test a localization system exclusively based on Machine
Vision for the use in AGVs. The x and y axis positioning as well as the determination of
orientation of the vehicle in all three
axes is carried out by a stereo camera based
Visual Odometry
approach and a supporting detection of artificial landmarks placed on the ceiling. Both methods
complement each other perfectly: while Visual Odometry bridges distances without landmarks,
drift caused by Visual Odometry is corrected by artificial landmarks. Test series have shown that
the localization error falls below ±20mm if the distance between camera and landmark does not
exceed 4500mm. Also the inclination of the vehicle is equalized. This localization system has
various advantages compared to well established methods: designing and installation efforts can
be reduced, while the flexibility for route changes can be increased compared to traditional
magnetic guidance systems. The interference immunity is higher compared to contour matching
methods due to the use of absolute reference points placed on the ceiling. The proposed system is
not suitable for
use in
halls because
the
distance between camera
and ceiling landmarks should not
exceed 4500mm. Nevertheless, this localization system is an interesting alternative to well
established methods primarily
for the
use in
the public sector, e.g. hospitals or libraries.
169
Proceedings
OAGM & ARW Joint Workshop 2016 on "Computer Vision and Robotics“
- Titel
- Proceedings
- Untertitel
- OAGM & ARW Joint Workshop 2016 on "Computer Vision and Robotics“
- Autoren
- Peter M. Roth
- Kurt Niel
- Verlag
- Verlag der Technischen Universität Graz
- Ort
- Wels
- Datum
- 2017
- Sprache
- englisch
- Lizenz
- CC BY 4.0
- ISBN
- 978-3-85125-527-0
- Abmessungen
- 21.0 x 29.7 cm
- Seiten
- 248
- Schlagwörter
- Tagungsband
- Kategorien
- International
- Tagungsbände