Improving the Agility of Keyframe-Based SLAM

  • Georg Klein
  • David Murray
Part of the Lecture Notes in Computer Science book series (LNCS, volume 5303)


The ability to localise a camera moving in a previously unknown environment is desirable for a wide range of applications. In computer vision this problem is studied as monocular SLAM. Recent years have seen improvements to the usability and scalability of monocular SLAM systems to the point that they may soon find uses outside of laboratory conditions. However, the robustness of these systems to rapid camera motions (we refer to this quality as agility) still lags behind that of tracking systems which use known object models. In this paper we attempt to remedy this. We present two approaches to improving the agility of a keyframe-based SLAM system: Firstly, we add edge features to the map and exploit their resilience to motion blur to improve tracking under fast motion. Secondly, we implement a very simple inter-frame rotation estimator to aid tracking when the camera is rapidly panning – and demonstrate that this method also enables a trivially simple yet effective relocalisation method. Results show that a SLAM system combining points, edge features and motion initialisation allows highly agile tracking at a moderate increase in processing time.


Point Feature Augmented Reality Coordinate Frame Motion Blur Edge Feature 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Supplementary material


  1. 1.
    Davison, A., Reid, I., Molton, N.D., Stasse, O.: MonoSLAM: Real-time single camera SLAM. IEEE Trans. Pattern Analysis and Machine Intelligence 29, 1052–1067 (2007)CrossRefGoogle Scholar
  2. 2.
    Chekhlov, D., Pupilli, M., Mayol-Cuevas, W., Calway, A.: Real-time and robust monocular SLAM using predictive multi-resolution descriptors. In: Proc 2nd International Symposium on Visual Computing (November 2006)Google Scholar
  3. 3.
    Lowe, D.: Distinctive image features from scale-invariant keypoints. International Journal of Computer Vision 60(2), 91–100 (2004)CrossRefGoogle Scholar
  4. 4.
    Williams, B., Klein, G., Reid, I.: Real-time SLAM relocalisation. In: Proc 11th IEEE International Conference on Computer Vision (ICCV 2007), Rio de Janeiro (October 2007)Google Scholar
  5. 5.
    Eade, E., Drummond, T.: Scalable monocular SLAM. In: Proc. IEEE Intl. Conference on Computer Vision and Pattern Recognition (CVPR 2006), New York, pp. 469–476 (2006)Google Scholar
  6. 6.
    Klein, G., Murray, D.: Parallel tracking and mapping for small AR workspaces. In: Proc Intl. Symposium on Mixed and Augmented Reality (ISMAR 2007), Nara (November 2007)Google Scholar
  7. 7.
    Rosten, E., Drummond, T.: Machine learning for high-speed corner detection. In: Leonardis, A., Bischof, H., Pinz, A. (eds.) ECCV 2006. LNCS, vol. 3951. Springer, Heidelberg (2006)Google Scholar
  8. 8.
    Klein, G., Drummond, T.: Tightly integrated sensor fusion for robust visual tracking. In: Proc. British Machine Vision Conference (BMVC 2002), Cardiff, pp. 787–796 (September 2002)Google Scholar
  9. 9.
    Smith, P., Reid, I., Davison, A.: Real-time monocular SLAM with straight lines. In: Proc British Machine Vision Conference (BMVC 2006), Edinburgh (September 2006)Google Scholar
  10. 10.
    Eade, E., Drummond, T.: Edge landmarks in monocular SLAM. In: Proc. British Machine Vision Conference (BMVC 2006), Edinburg (September 2006)Google Scholar
  11. 11.
    Viéville, T., Faugeras, O.D.: Feedforward recovery of motion and structure from a sequence of 2d-lines matches. In: Proc. 3rd Int. Conf. on Computer Vision, pp. 517–520 (1990)Google Scholar
  12. 12.
    Faugeras, O.D., Lustaman, F., Toscani, G.: Motion and structure from point and line matches. In: Proc. 1st Int. Conf. on Computer Vision, London, pp. 25–33 (1987)Google Scholar
  13. 13.
    Taylor, C.J., Kriegman, D.J.: Structure and motion from line segments in multiple images. IEEE Transactions on Pattern Analysis and Machine Intelligence 17(11), 1021–1032 (1995)CrossRefGoogle Scholar
  14. 14.
    Bartoli, A., Sturm, P.: Structure from motion using lines: Representation, triangulation and bundle adjustment. Computer Vision and Image Understanding 100(3), 416–441 (2005)CrossRefGoogle Scholar
  15. 15.
    Shashua, A.: Trilinearity in visual recognition by alignment. In: Proc. 3rd European Conf. on Computer Vision, Stockholm, May 1994, pp. 479–484. Springer, Heidelberg (1994)Google Scholar
  16. 16.
    Devernay, F., Faugeras, O.D.: Straight lines have to be straight. Machine Vision and Applications 13(1), 14–24 (2001)CrossRefGoogle Scholar
  17. 17.
    Harris, C.: Tracking with rigid models. In: Blake, A. (ed.) Active Vision, MIT Press, Cambridge (1992)Google Scholar
  18. 18.
    Canny, J.: A computational approach to edge detection. IEEE Transactions on Pattern Analysis and Machine Intelligence (PAMI) 8(6), 679–698 (1986)CrossRefGoogle Scholar
  19. 19.
    Benhimane, S., Malis, E.: Homography-based 2d visual tracking and servoing. Special Joint Issue on Robotics and Vision. Journal of Robotics Research 26(7), 661–676 (2007)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2008

Authors and Affiliations

  • Georg Klein
    • 1
  • David Murray
    • 1
  1. 1.Active Vision LaboratoryUniversity of OxfordUK

Personalised recommendations