Advertisement

Probabilistic Model-Based Background Subtraction

  • Volker Krüger
  • Jakob Anderson
  • Thomas Prehn
Part of the Lecture Notes in Computer Science book series (LNCS, volume 3540)

Abstract

Usually, background subtraction is approached as a pixel-based process, and the output is (a possibly thresholded) image where each pixel reflects, independent from its neighboring pixels, the likelihood of itself belonging to a foreground object. What is neglected for better output is the correlation between pixels. In this paper we introduce a model-based background subtraction approach which facilitates prior knowledge of pixel correlations for clearer and better results. Model knowledge is being learned from good training video data, the data is stored for fast access in a hierarchical manner. Bayesian propagation over time is used for proper model selection and tracking during model-based background subtraction. Bayes propagation is attractive in our application as it allows to deal with uncertainties during tracking. We have tested our approach on suitable outdoor video data.

Keywords

Model Knowledge Foreground Object Gait Recognition Sequential Importance Sampling Markov Transition Matrix 
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.

References

  1. 1.
    Comaniciu, D., Ramesh, V., Meer, P.: Real-time tracking of non-rigid objects using mean shift. In: Proc. IEEE Conf. on Computer Vision and Pattern Recognition, Hilton Head Island, SC, June 13-15, vol. 2, pp. 142–149 (2000)Google Scholar
  2. 2.
    Doucet, A., Godsill, S., Andrieu, C.: On sequential monte carlo sampling methods for bayesian filtering. Statistics and Computing 10, 197–209 (2000)CrossRefGoogle Scholar
  3. 3.
    Elgammal, A., Davis, L.: Probabilistic framework for segmenting people under occlusion. In: ICCV 2001 (2001)Google Scholar
  4. 4.
    Gavrila, D., Philomin, V.: Real-time object detection for ”smart” vehicles. In: Proc. Int. Conf. on Computer Vision, Korfu, Greece, pp. 87–93 (1999)Google Scholar
  5. 5.
    Haritaoglu, I., Harwood, D., Davis, L.: W4s: A real-time system for detection and tracking people in 2.5 D. In: Proc. European Conf. on Computer Vision, Freiburg, Germany, June 1-5 (1998)Google Scholar
  6. 6.
    Horprasert, T., Harwood, D., Davis, L.S.: A statistical approach for real-time robust background subtraction and shadow detection. In: Proceedings of IEEE ICCV 1999 FRAME-RATE Workshop (1999)Google Scholar
  7. 7.
    Isard, M., Blake, A.: Condensation – conditional density propagation for visual tracking. Int. J. of Computer Vision (1998)Google Scholar
  8. 8.
    Isard, M., Blake, A.: Condensation – conditional density propagation for visual tracking. Int. J. of Computer Vision 29, 5–28 (1998)CrossRefGoogle Scholar
  9. 9.
    Ivanov, Y.A., Bobick, A.F., Liu, J.: Fast lighting independent background subtraction. Int. J. of Computer Vision 37(2), 199–207 (2000)zbMATHCrossRefGoogle Scholar
  10. 10.
    Kale, A., Sundaresan, A., Rjagopalan, A.N., Cuntoor, N., Chowdhury, A.R., Krüger, V., Chellappa, R.: Identification of humans using gait. IEEE Trans. Image Processing 9, 1163–1173 (2004)CrossRefGoogle Scholar
  11. 11.
    Kitagawa, G.: Monta carlo filter and smoother for non-gaussian nonlinear state space models. J. Computational and Graphical Statistics 5, 1–25 (1996)CrossRefMathSciNetGoogle Scholar
  12. 12.
    Krueger, V., Zhou, S.: Exemplar-based face recognition from video. In: Proc. European Conf. on Computer Vision, Copenhagen, Denmark, June 27-31 (2002)Google Scholar
  13. 13.
    Liu, J.S., Chen, R.: Sequential monte carlo for dynamic systems. Journal of the American Statistical Association 93, 1031–1041 (1998)Google Scholar
  14. 14.
    Toyama, K., Blake, A.: Probabilistic tracking in a metric space. In: Proc. Int. Conf. on Computer Vision, Vancouver, Canada, July 9-12, vol. 2, pp. 50–59 (2001)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2005

Authors and Affiliations

  • Volker Krüger
    • 1
  • Jakob Anderson
    • 2
  • Thomas Prehn
    • 2
  1. 1.Aalborg Media LabAalborg UniversityCopenhagen, Ballerup
  2. 2.Aalborg University EsbjergEsbjergDenmark

Personalised recommendations