Incorporating building information to globalize and robustify grid-based indoor SLAM

Beitrag bei einer Tagung
(Konferenzbeitrag)


Details zur Publikation

Autor(en): Hiller M, Particke F, Patino-Studencki L, Thielecke J
Jahr der Veröffentlichung: 2017
Tagungsband: 2017 17th International Conference on Control, Automation and Systems (ICCAS)
Seitenbereich: 1872-1877


Abstract


Navigation is one of the key topics in the field of mobile robotics. In many areas of application like Industry 4.0 or fully automated parking, additional external sensors are available as well as pre-existing knowledge about the general building structure. This information is of significant advantage, especially in situations where the perceptive field of the mobile platform is occluded. One method to gain the required information is the probabilistic concept of simultaneous localization and mapping (SLAM). However, most existing approaches generally lack the possibility to incorporate preexisting and external information. In this paper, a solution to the SLAM problem based on a Rao-Blackwellized particle filter is adopted to provide efficient means for exploiting such data. We present an approach that directly incorporates environment knowledge using a context-aware environment model, while establishing reference to a global coordinate frame allowing for a straight-forward fusion with external information sources. The evaluation is performed on real-world data obtained by a mobile platform. The qualitative analysis shows significant improvement in map quality and robustness regarding short sensor outages or reduced perception rates. Establishing a uniform reference frame and reusing data, the proposed approach clearly extends the functional range of SLAM, demonstrating substantial advantages over existing methods.



FAU-Autoren / FAU-Herausgeber

Hiller, Markus
Professur für Informationstechnik (Schwerpunkt Ortsbestimmung und Navigation)
Particke, Florian
Professur für Informationstechnik (Schwerpunkt Ortsbestimmung und Navigation)
Patino-Studencki, Lucila
Professur für Informationstechnik (Schwerpunkt Ortsbestimmung und Navigation)
Thielecke, Jörn Prof. Dr.
Professur für Informationstechnik (Schwerpunkt Ortsbestimmung und Navigation)


Zitierweisen

APA:
Hiller, M., Particke, F., Patino-Studencki, L., & Thielecke, J. (2017). Incorporating building information to globalize and robustify grid-based indoor SLAM. In 2017 17th International Conference on Control, Automation and Systems (ICCAS) (pp. 1872-1877). Jeju, KR.

MLA:
Hiller, Markus, et al. "Incorporating building information to globalize and robustify grid-based indoor SLAM." Proceedings of the Control, Automation and Systems (ICCAS), 2017 17th International Conference on, Jeju 2017. 1872-1877.

BibTeX: 

Zuletzt aktualisiert 2018-18-10 um 21:20

Link teilen