Online Moving Camera Background Subtraction

  • Ali Elqursh
  • Ahmed Elgammal
Part of the Lecture Notes in Computer Science book series (LNCS, volume 7577)


Recently several methods for background subtraction from moving camera were proposed. They use bottom up cues to segment video frames into foreground and background regions. Due to this lack of explicit models, they can easily fail to detect a foreground object when such cues are ambiguous in certain parts of the video. This becomes even more challenging when videos need to be processed online. We present a method which enables learning of pixel based models for foreground and background regions and, in addition, segments each frame in an online framework. The method uses long term trajectories along with a Bayesian filtering framework to estimate motion and appearance models. We compare our method to previous approaches and show results on challenging video sequences.


Motion Vector Gaussian Mixture Model Background Subtraction Motion Model Appearance Model 
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.
    Belkin, M., Niyogi, P.: Laplacian Eigenmaps for Dimensionality Reduction and Data. Neural Computation 15, 1373–1396 (2003)zbMATHCrossRefGoogle Scholar
  2. 2.
    Boykov, Y., Funka-Lea, G.: Graph Cuts and Efficient N-D Image Segmentation. International Journal of Computer Vision 70(2), 109–131 (2006)CrossRefGoogle Scholar
  3. 3.
    Brox, T., Malik, J.: Object Segmentation by Long Term Analysis of Point Trajectories. In: Daniilidis, K., Maragos, P., Paragios, N. (eds.) ECCV 2010, Part V. LNCS, vol. 6315, pp. 282–295. Springer, Heidelberg (2010)CrossRefGoogle Scholar
  4. 4.
    Costeira, J., Kanade, T.: A multi-body factorization method for motion analysis. In: ICCV, pp. 1071–1076 (1995)Google Scholar
  5. 5.
    Elgammal, A., Duraiswami, R., Harwood, D., Davis, L.: Background and foreground modeling using nonparametric kernel density estimation for visual surveillance. Proceedings of the IEEE 90(7), 1151–1163 (2002)CrossRefGoogle Scholar
  6. 6.
    Elhamifar, E., Vidal, R.: Sparse subspace clustering. In: CVPR, pp. 2790–2797 (June 2009)Google Scholar
  7. 7.
    Irani, M., Rousso, B., Peleg, S.: Computing occluding and transparent motions. International Journal of Computer Vision 12(1), 5–16 (1994)CrossRefGoogle Scholar
  8. 8.
    Kanatani, K.: Motion segmentation by subspace separation and model selection. In: ICCV, vol. 2, pp. 586–591 (2001)Google Scholar
  9. 9.
    Kwak, S., Lim, T., Nam, W., Han, B., Hee, J.: Generalized Background Subtraction Based on Hybrid Inference by Belief Propagation and Bayesian Filtering. In: ICCV (2011)Google Scholar
  10. 10.
    Mittal, A., Paragios, N.: Motion-based background subtraction using adaptive kernel density estimation. In: CVPR, vol. 2, pp. 302–309. IEEE (2004)Google Scholar
  11. 11.
    Ochs, P., Brox, T.: Object Segmentation in Video: A Hierarchical Variational Approach for Turning Point Trajectories into Dense Regions. In: ICCV (2011)Google Scholar
  12. 12.
    Pawan Kumar, M., Torr, P.H.S., Zisserman, A.: Learning Layered Motion Segmentations of Video. IJCV 76(3), 301–319 (2007)CrossRefGoogle Scholar
  13. 13.
    Rao, S.R., Tron, R.: Motion Segmentation via Robust Subspace Separation in the Presence of Outlying, Incomplete, or Corrupted Trajectories. In: CVPR (2008)Google Scholar
  14. 14.
    Rowe, S., Blake, A.: Statistical mosaics for tracking. Image and Vision Computing 14(8), 549–564 (1996)CrossRefGoogle Scholar
  15. 15.
    Sheikh, Y., Javed, O., Kanade, T.: Background Subtraction for Freely moving cameras. In: ICCV (2009)Google Scholar
  16. 16.
    Shi, J.: Motion segmentation and tracking using normalized cuts. In: ICCV, vol. 32(10), pp. 1832–1845 (October 2002)Google Scholar
  17. 17.
    Stauffer, C.: Learning patterns of activity using real-time tracking. PAMI 22(8), 747–757 (2000)CrossRefGoogle Scholar
  18. 18.
    Sundaram, N., Brox, T., Keutzer, K.: Dense Point Trajectories by GPU-Accelerated Large Displacement Optical Flow. In: Daniilidis, K., Maragos, P., Paragios, N. (eds.) ECCV 2010, Part I. LNCS, vol. 6311, pp. 438–451. Springer, Heidelberg (2010)CrossRefGoogle Scholar
  19. 19.
    Tao, H., Sawhney, H.S., Kumar, R.: Object Tracking with Bayesian Estimation of Dynamic Layer Representations. PAMI 24(1), 75–89 (2002)CrossRefGoogle Scholar
  20. 20.
    Torr, P.H.S.: Outlier detection and motion segmentation. PhD thesis, University of Oxford (1995)Google Scholar
  21. 21.
    Tron, R., Vidal, R.: A Benchmark for the Comparison of 3-D Motion Segmentation Algorithms. In: CVPR (2007)Google Scholar
  22. 22.
    Weiss, Y.: Smoothness in layers: Motion segmentation using nonparametric mixture estimation. In: CVPR, pp. 520–526 (1997)Google Scholar
  23. 23.
    Weiss, Y., Freeman, W.T.: Correctness of belief propagation in Gaussian graphical models of arbitrary topology. Neural Computation 13(10), 2173–2200 (2001)zbMATHCrossRefGoogle Scholar
  24. 24.
    Wren, C., Azarbayejani, A., Darrell, T., Pentland, A.: Pfinder: real-time tracking of the human body. PAMI 19(7), 780–785 (1997)CrossRefGoogle Scholar
  25. 25.
    Xiao, J.: Accurate Motion Layer Segmentation and Matting. In: CVPR, pp. 698–703 (2005)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2012

Authors and Affiliations

  • Ali Elqursh
    • 1
  • Ahmed Elgammal
    • 1
  1. 1.Rutgers UniversityUSA

Personalised recommendations