UNSUPERVISED LEARNING FOR TEMPORAL SEARCH SPACE REDUCTION IN THREE-DIMENSIONAL SCENE RECOVERY

Tom Warsop, Sameer Singh

Abstract

Methods for three-dimensional scene recovery traverse scene spaces (typically along epipolar lines) to compute two-dimensional image feature correspondences. These methods ignore potentially useful temporal information presented by previously processed frames, which can be used to decrease search space traversal. In this work, we present a general framework which models relationships between image information and recovered scene information specifically for the purpose of improving efficiency of three-dimensional scene recovery. We further present three different methods implementing this framework using either a naive Nearest Neighbour approach or a more sophisticated collection of associated Gaussians. Whilst all three methods provide a decrease in search space traversal, it is the Gaussian-based method which performs best, as the other methods are subject to the (demonstrated) unwanted behaviours of convergence and oscillation.

References

  1. Bay, H., Ess, A., Tuytelaars, T., and Gool, L. V. (2008). Surf: Speeded up robust features. Computer Vision and Image Understanding (CVIU), 110(3):346-359.
  2. Chekhlov, D. and Mayol-Cuevas, W. (2008). Appearance based indexing for relocalisation in real-time visual slam. In 19th British Machine Vision Conference, pages 363-372.
  3. Davison, A. J. (2003). Real-time simulataneous localization and mapping with a single camera. In Proc. International Conference on Computer Vision, pages 1403- 1411.
  4. Davison, A. J., Reid, I. D., Molton, N. D., and Stasse, O. (2007). Monoslam: Real-time single camera slam. In IEEE Transactions on Pattern Analysis and Machine Intelligence, volume 29, pages 1-15.
  5. Fabbri, R. and Kimia, B. (2010). 3d curve sketch: Flexible curve-based stereo reconstruction ad calibration. In 2010 IEEE Conference on Computer Vision and Pattern Recognition, pages 1538-1545.
  6. Favaro, P., Jin, H., and Soatto, S. (2003). A semi-direct approach to structure from motion. In The Visual Computer, volume 19, pages 377-384.
  7. Grinberg, M., Ohr, F., and ad J. Beyerer, D. W. (2010). Feature-based probabilistic data association and tracking. In The 7th International Workshop on Intelligent Transportation (WIT2010), pages 29-34.
  8. Klein, G. and Murray, D. (2007). Parallel tracking and mapping for small ar workspaces. In Proceedings and the Sixth IEEE and ACM International Symposium on Mixed and Augmented Reailty (ISMAR07).
  9. Li, J., Li, ., Chen, Y., Xu, L., and Zhang, Y. (2010). Bundled depth-map merging for multi-view stereo. In 2010 IEEE Conference on Computer Vision and Pattern Recognition, pages 2769-2776.
  10. Li, P., Farin, D., Gunnewiek, R. K., and de With, P. H. N. (2006). On creating depth maps from monoscopic video using structure frm motion. In 27th Symposium on Information Theory, pages 508-515.
  11. Warsop, T. E. and Singh, S. (2010). Robust threedimensional scene recovery from monocular image pairs. In 9th IEEE International Conference on Cybernetic Intelligent Systems 2010 (CIS 2010), pages 112-117.
  12. Zhang, G., Dong, Z., Jia, J., Wong, T.-T., and Bao, H. (2010). Efficient non-consecutive feature tracking for structure-from-motion. In ECCV 2010, pages 422- 435.
  13. Zhang, G., Jia, J., Wong, T.-T., and Bao, H. (2009). Consistent depth maps recovery from a video sequence. In IEEE Transactions on Pattern Analysis and Machine Intelligence, volume 21, pages 974-988.
Download


Paper Citation


in Harvard Style

Warsop T. and Singh S. (2011). UNSUPERVISED LEARNING FOR TEMPORAL SEARCH SPACE REDUCTION IN THREE-DIMENSIONAL SCENE RECOVERY . In Proceedings of the International Conference on Computer Vision Theory and Applications - Volume 1: VISAPP, (VISIGRAPP 2011) ISBN 978-989-8425-47-8, pages 549-554. DOI: 10.5220/0003308405490554


in Bibtex Style

@conference{visapp11,
author={Tom Warsop and Sameer Singh},
title={UNSUPERVISED LEARNING FOR TEMPORAL SEARCH SPACE REDUCTION IN THREE-DIMENSIONAL SCENE RECOVERY},
booktitle={Proceedings of the International Conference on Computer Vision Theory and Applications - Volume 1: VISAPP, (VISIGRAPP 2011)},
year={2011},
pages={549-554},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0003308405490554},
isbn={978-989-8425-47-8},
}


in EndNote Style

TY - CONF
JO - Proceedings of the International Conference on Computer Vision Theory and Applications - Volume 1: VISAPP, (VISIGRAPP 2011)
TI - UNSUPERVISED LEARNING FOR TEMPORAL SEARCH SPACE REDUCTION IN THREE-DIMENSIONAL SCENE RECOVERY
SN - 978-989-8425-47-8
AU - Warsop T.
AU - Singh S.
PY - 2011
SP - 549
EP - 554
DO - 10.5220/0003308405490554