Combining multiple motion estimates for vehicle tracking

  • Sylvia Gil
  • Ruggero Milanese
  • Thierry Pun
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 1065)


In this paper, the problem of combining estimates provided by multiple models is considered, with application to vehicle tracking. Two tracking systems, based on the bounding-box and on the 2-D pattern of the targets, provide individual motion parameters estimates to the combining method, which in turn produces a global estimate. Two methods are proposed to combine the estimates of these tracking systems: one is based on their covariance matrix, while the other one employs a Kalman filter model. Results are provided on three image sequences taken under different viewpoints, weather conditions and varying vehicle/road contrasts. Two evaluations are made. First, the performances of individual and global estimates are compared. Second, the two global estimates are compared and the superiority of the second method is assessed over the first one.


Kalman Filter Image Sequence Tracking System Position Error Global Estimate 
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.


  1. 1.
    V. Tresp, M. Taniguchi, “Combining Estimators Using Non-Constant Weighting Functions”, to appear in Proc. of the Neural Information Processing Systems, 1995.Google Scholar
  2. 2.
    D. Koller, K. Daniilidis, H.H. Nagel, Model-Based Object Tracking in Monocular Image Sequences of Road Traffic Scenes, Int. Jour. of Computer Vision, Vol. 3, pp. 257–281, 1993.Google Scholar
  3. 3.
    A. Gelb, Applied Optimal Estimation, The MIT Press, MA, and London, UK, 1974.Google Scholar
  4. 4.
    K. Baker, G. D. Sullivan, Performance Assessment of Model-based Tracking, Proc. of the IEEE Workshop on Applications of Computer Vision, pp. 28–35, Palm Springs, CA, 1992.Google Scholar
  5. 5.
    F. Meyer, P. Bouthémy, Region-Based Tracking in an Image Sequence, Proceedings of the European Conference on Computer Vision, pp. 476–484, S. Margarita-Ligure, Italy, 1992.Google Scholar
  6. 6.
    A. Blake, R. Curwen, A. Zisserman, A Framework for Spatiotemporal Control in the Tracking of Visual Contours, Int. Journal of Computer Vision, Vol. 11, pp. 127–147, 1993.Google Scholar
  7. 7.
    D. Koller, J. Weber, J. Malik, Robust Multiple Car Tracking with Occlusion Reasoning, Proceedings of the Third European Conference on Computer Vision, Vol. 1, pp. 189–199, Stockholm, Sweden, 1994.Google Scholar
  8. 8.
    O. Faugeras, Three-Dimensional Computer Vision: A Geometric Viewpoint, The MIT Press, London, UK, 1993.Google Scholar
  9. 9.
    J. Shi, C. Tomasi, Good Features to Track, IEEE Conf. on Computer Vision and Pattern Recognition, pp. 593–600, Seattle, USA, 1994.Google Scholar
  10. 10.
    S. Gil, R. Milanese, T. Pun, “Comparing Features for Target Tracking in Traffic Scenes”, to appear in Pattern Recognition.Google Scholar
  11. 11.
    D. Geiger, J.A. Vlontzos, Matching Elastic contours, IEEE Conf. on Computer Vision and Pattern Recognition, pp. 602–604, New York, June 1993.Google Scholar
  12. 12.
    International Journal of Forecasting, special issue on combining forecasts, Vol. 4(4), 1989.Google Scholar
  13. 13.
    F. Leymarie, D. Levine, Tracking Deformable Objects in the Plane Using an Active Contour Model, IEEE Trans. on Pattern Analysis and Machine Intelligence, Vol. 15, 1993.Google Scholar
  14. 14.
    F. P. Preparata, M. I. Shamos, Computational Geometry, Springer-Verlag, 1985.Google Scholar
  15. 15.
    S. Gil, T. Pun, Non-linear Multiresolution Relaxation for Alerting, Proceedings of the European Conference on Circuit Theory and Design, pp. 1639–1644, Davos, CH, 1993.Google Scholar
  16. 16.
    C.V. Granger, “Combining forecasts — twenty years later”, Journal of Forecasting, Vol. 8, pp. 167–173, 1989.Google Scholar
  17. 17.
    M.P. Perrone and L.N. Cooper, “When networks disagree: Ensemble methods for hybrid Neural Networks”, in “Neural Networks for Speech and Processing”, Editor R.J. Mammone, Chapman-Hall, 1993.Google Scholar
  18. 18.
    R. Meir, “Bias, variance and the combination of estimators; the case of linear least squares”, Preprint, Technion, Heifa, Israel, 1994.Google Scholar
  19. 19.
    S. Hashem, “Optimal Linear Combinations of Neural Networks”, Ph.D. Thesis, Purdue University, December 1993.Google Scholar
  20. 20.
    A. Krogh, J. Vedelsby, “Neural Network Ensemble, Cross Validation and Active Learning”, to appear in Proc. of the Neural Information Processing Systems, 1995.Google Scholar
  21. 21.
    Y. Bar-Shalom, “Multitarget Multisensor Tracking: Advanced Applications”, Artech, 1990.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1996

Authors and Affiliations

  • Sylvia Gil
    • 1
  • Ruggero Milanese
    • 1
  • Thierry Pun
    • 1
  1. 1.Computer Science DepartmentUniversity of GenevaGenÊve-4Switzerland

Personalised recommendations