Improving SLAM by Exploiting Building Information from Publicly Available Maps and Localization Priors

Original Article
  • 115 Downloads

Abstract

Maps are needed for a wide range of applications. In the context of mobile robotics, the map learning problem under uncertainty is often referred to as the simultaneous localization and mapping problem. In this paper, we aim at exploiting already available information such as OpenStreetMap data within the Simultaneous Localization and Mapping (SLAM) problem. We achieve this by relating the information about buildings with the perceptions of the robot and generate constraints for the pose graph-based formulation of the SLAM problem. In addition to that, we present a way to select target locations for the robot so that by going there, the robot can expect to reduce its own pose uncertainty. This localizability information is generated directly from OpenStreetMap data and supports active localization. We implemented and evaluated our approach using real-world data taken in urban environments. Our experiments suggest that we are able to relate the newly built maps with information from OpenStreetMap with the laser range finder data from the robot and in this way improve the map quality. The extension to graph-based SLAM provides better aligned maps and adds only a marginal computational overhead. Furthermore, we illustrate that the localizability information is useful to evaluate the ability to localize the robot given a trajectory.

Keywords

SLAM OpenStreetMap Alignment Active localization 

Zusammenfassung

Verbesserung von SLAM durch öffentlich verfügbare Gebäudedaten. Nahezu alle Navigationssysteme benötigen Karten der Umgebung. Das gleichzeitige Erstellen und Nutzen solcher Karten spielt eine zentrale Rolle in der Roboternavigation und wird oft als Simultaneous Localization and Mapping oder SLAM Problem bezeichnet. Nahezu alle gängigen SLAM Systeme ignorieren allerdings Hintergrundwissen oder Resourcen aus dem Netz. In diesem Papier präsentieren wir ein Verfahren, welches Daten von OpenStreetMap während des Kartenaufbaus nutzen kann. Dazu erweitern wir die klassische Pose Graph Formulierung des SLAM Problems und integrieren zusätzliche Abhängigkeiten zwischen Aufnahmeposen und existierendem Kartenmaterial. Darüber hinaus können wir schätzen, welche Regionen der Karte sich zur Positionsbestimmung besonders eignen und somit die Trajektorien des Roboters positiv bezüglich der erwarteten Lokalisierbarkeit bewerten. Wir haben unseren Ansatz auf zwei realen Robotersystemen implementiert und evaluiert. Wie unsere Experimente zeigen, verbessert unser Verfahren die resultierenden Karten ohne dabei den Rechenaufwand substanziell zu erhöhen.

References

  1. Agarwal P, Tipaldi G, Spinello L, Stachniss C, Burgard W (2013) Robust map optimization using dynamic covariance scaling. In: Proceedings of the IEEE international conference on robotics & automation (ICRA), pp 62–69Google Scholar
  2. Aloimonos J, Weiss I, Bandyopadhyay A (1988) Active vision. Int J Comput Vis 1(4):333–356CrossRefGoogle Scholar
  3. Bailey T, Durrant-Whyte H (2006a) Simultaneous localisation and mapping (SLAM): part I. Robot Autom Mag 13(2):99–110Google Scholar
  4. Bailey T, Durrant-Whyte H (2006b) Simultaneous localisation and mapping (SLAM): part II. Robot Autom Mag 13(3):108–117Google Scholar
  5. Bajcsy R (1988) Active perception. Proc IEEE 76(8):966–1005CrossRefGoogle Scholar
  6. Bengtsson O, Baerveldt A-J (2003) Robot localization based on scan-matching estimating the covariance matrix for the IDC algorithm. Robot Autonom Syst 44(1):29–40CrossRefGoogle Scholar
  7. Besl P, McKay ND (1992) A method for registration of 3-d shapes. IEEE Trans Pattern Anal Mach Intell 14(2):239–256CrossRefGoogle Scholar
  8. Bogoslavskyi I, Mazuran M, Stachniss C (2016) Robust homing for autonomous robots. In: Proceedings of the IEEE international conference on robotics & automation (ICRA), pp 2550–2556Google Scholar
  9. Brubaker M, Geiger A, Urtasun R (2016) Map-based probabilistic visual self-localization. IEEE Trans Pattern Anal Mach Intell 38(4):652–665CrossRefGoogle Scholar
  10. Chen S, Li Y, Kwok NM (2011) Active vision in robotic systems: a survey of recent developments. Int J Robot Res 30(11):1343–1377Google Scholar
  11. Dequaire J, Tong CH, Churchill W, Posner I (2016) Off the beaten track: predicting localisation performance in visual teach and repeat. In: Proceedings of the IEEE international conference on robotics & automation (ICRA), pp 795–800Google Scholar
  12. Fischer A, Kolbe TH, Lang F, Cremers AB, Förstner W, Plümer L, Steinhage V (1998) Extracting buildings from aerial images using hierarchical aggregation in 2d and 3d. Comput Vis Image Understand 72(2):185–203CrossRefGoogle Scholar
  13. Floros G, van der Zander B, Leibe B (2013) Openstreetslam: global vehicle localization using openstreetmaps. In: Proceedings of the IEEE international conference on robotics & automation (ICRA), pp 1054–1059Google Scholar
  14. Förstner W, Wrobel B (2016) Photogrammetric computer vision. In: Chapter robust estimation and outlier detection. Springer, Berlin, pp 141–159Google Scholar
  15. Furgale P, Barfoot TD (2010) Visual teach and repeat for long-range rover autonomy. J Field Robot 27(5):534–560Google Scholar
  16. Gerke M (2011) Using horizontal and vertical building structure to constrain indirect sensor orientation. ISPRS J Photogramm Remote Sens 66(3):307–316CrossRefGoogle Scholar
  17. Grisetti G, Kümmerle R, Stachniss C, Burgard W (2010a) A tutorial on graph-based SLAM. IEEE Trans Intell Transp Syst Mag 2:31–43Google Scholar
  18. Grisetti G, Kümmerle R, Stachniss C, Frese U, Hertzberg C (2010b) Hierarchical optimization on manifolds for online 2D and 3D mapping. In: Proceedings of the IEEE international conference on robotics & automation (ICRA), pp 273–278Google Scholar
  19. Grisetti G, Stachniss C, Burgard W (2009) Non-linear constraint network optimization for efficient map learning. IEEE Trans Intell Transp Syst 10:428–439CrossRefGoogle Scholar
  20. Gutmann JS, Konolige K (2000) Incremental mapping of large cyclic environments. In: Proceedings of the IEEE international symposium on computational intelligence in robotics and automation (CIRA), pp 318–325Google Scholar
  21. Haklay M, Weber P (2008) Openstreetmap: user-generated street maps. Pervasive computing. IEEE 7(4):12–18Google Scholar
  22. Hentschel M, Wagner B (2010) Autonomous robot navigation based on openstreetmap geodata. In: 2010 13th international IEEE conference on intelligent transportation systems (ITSC). IEEE, New York, pp 1645–1650Google Scholar
  23. Hentschel M, Wulf O, Wagner B (2008) A gps and laser-based localization for urban and non-urban outdoor environments. In: Proceedings of the IEEE/RSJ international conference on intelligent robots and systems (IROS), pp 149–154Google Scholar
  24. Huber M, Schickler W, Hinz S, Baumgartner A (2003) Fusion of lidar data and aerial imagery for automatic reconstruction of building surfaces. In: 2nd GRSS/ISPRS joint workshop on remote sensing and data fusion over urban areas, 2003, pp 82–86Google Scholar
  25. Kim A, Eustice RM (2015) Active visual slam for robotic area coverage: theory and experiment. Int J Robot Res 34(4–5):457–475CrossRefGoogle Scholar
  26. Konolige K, Grisetti G, Kümmerle R, Burgard W, Limketkai B, Vincent R (2010) Sparse pose adjustment for 2d mapping. In: Proceedings of the IEEE/RSJ international conference on intelligent robots and systems (IROS), pp 22–29Google Scholar
  27. Kümmerle R, Grisetti G, Strasdat H, Konolige K, Burgard W (2011a) g2o: a general framework for graph optimization. In: Proceedings of the IEEE international conference on robotics & automation (ICRA), pp 3607–3613Google Scholar
  28. Kümmerle R, Steder B, Dornhege C, Kleiner A, Grisetti G, Burgard W (2011b) Large scale graph-based slam using aerial images as prior information. Autonom Robots 30(1):25–39CrossRefGoogle Scholar
  29. Latif Y, Cadena C, Neira J (2012) Robust loop closing over time. In: Proceedings of robotics: science and systems (RSS)Google Scholar
  30. Lu F, Milios E (1997) Globally consistent range scan alignment for environment mapping. Autonom Robots 4:333–349CrossRefGoogle Scholar
  31. Olson E, Leonard J, Teller S (2006) Fast iterative optimization of pose graphs with poor initial estimates. In: Proceedings of the IEEE internationa conference on robotics & automation (ICRA), pp 2262–2269Google Scholar
  32. Pink O, Moosmann F, Bachmann A (2009) Visual features for vehicle localization and ego-motion estimation. In: IEEE intelligent vehicles symposium, pp 254–260Google Scholar
  33. Roy N, Burgard W, Fox D, Thrun S (1998) Coastal navigation – robot motion with uncertainty. In: Proceedings of the AAAI fall symposium: planning with POMDPs, Stanford, CA, USAGoogle Scholar
  34. Ruchti P, Steder B, Ruhnke M, Burgard W (2015) Localization on openstreetmap data using a 3d laser scanner. In: Proceedings of the IEEE international conference on robotics & automation (ICRA), pp 5260–5265Google Scholar
  35. Salas-Moreno RF, Newcombe RA, Strasdat H, Kelly PH, Davison AJ (2013) Slam++: simultaneous localisation and mapping at the level of objects. In: Proceedings of the IEEE conference on computer vision and pattern recognition, pp 1352–1359Google Scholar
  36. Stachniss C (2017) Simultaneous Localization and Mapping. In: Springer Handbuch der Geodaesie. Springer, Berlin, pp 1–29Google Scholar
  37. Sünderhauf N, Protzel P (2012) Switchable constraints for robust pose graph slam. In: Proceedings of the IEEE/RSJ international conference on intelligent robots and systems (IROS), pp 1879–1884Google Scholar
  38. Thrun S, Montemerlo M (2006) The graph SLAM algorithm with applications to large-scale mapping of urban structures. Int J Robot Res 25(5–6):403Google Scholar
  39. Unger J, Rottensteiner F, Heipke C (2016) Integration of a generalised building model into the pose estimation of uas images. ISPRS—international archives of the photogrammetry. Remote Sens Spatial Inf Sci XLI-B1:1057–1064Google Scholar
  40. Velez J, Hemann G, Huang AS, Posner I, Roy N (2011) Planning to perceive: exploiting mobility for robust object detection. In: International conference on automated planning and scheduling (ICAPS), pp 266–273Google Scholar
  41. Vysotska O, Stachniss C (2016) Exploiting building information from publicly available maps in graph-based slam. In: Proceedings of the IEEE/RSJ international conference on intelligent robots and systems (IROS), pp 4511–4516Google Scholar
  42. Wulf O, Arras KO, Christensen HI, Wagner B (2004) 2d mapping of cluttered indoor environments by means of 3d perception. In: Proceedings of the IEEE international conference on robotics & automation (ICRA), pp 4204–4209Google Scholar

Copyright information

© Deutsche Gesellschaft für Photogrammetrie, Fernerkundung und Geoinformation (DGPF) e.V. 2017

Authors and Affiliations

  1. 1.Institute of Geodesy and GeoinformationUniversity of BonnBonnGermany

Personalised recommendations