Robust Estimation of the Optical Flow Based on VQ-BF

  • Manuel Graña
Part of the Studies in Fuzziness and Soft Computing book series (STUDFUZZ, volume 109)


To increase the robustness of optical flow computation in the context of mobile robotics, we introduce an image filtering process based on the codebook computed by Vector Quantization (VQ) algorithms, which usually are used for compression and codification purposes. The Self Organizing Map is used to compute adaptively the vector quantizers of color imaged sequences. The codebook computed for each image in the sequence is then used as a smoothing filter, the VQ Bayesian Filter (VQ-BF), for preprocessing images in the sequence. The filtered images are the basis for the computation of the optical flow via pixel and region correlation algorithms. The pixel correlation gives a good estimation of the optical flow at the image edges, whereas the region correlation gives a robust and dense estimation of the optical flow.


Displacement Field Optical Flow Vector Quantization Image Block Flow Estimation 
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.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    M.J. Black, Y. Yacoob, S.X. Ju (1997) “Recognizing human motion using parameterized models of optical flow” in “Motion Based Recognition” M. Shah, R. Jain (eds) Dordrecht: Kluwer Acad. Pub. pp. 245–269Google Scholar
  2. 2.
    T. Camus (1997) “Real time quantized optical flow” Real Time Imaging 3 (2): 71–80CrossRefGoogle Scholar
  3. 3.
    K. Chaudhury, R. Mehrotra (1995) “A trajectory based computational model for optical flow estimation” IEEE trans Rob. Aut. 11 (5) pp 733–741CrossRefGoogle Scholar
  4. 4.
    Cosman P.C, Oehler K.L., Riskin E.A., Gray R.M. (1993) “Using Vector Quantization for Image Processing” IEEE Proceedings 81 (4): 1326–1341CrossRefGoogle Scholar
  5. 5.
    I. Essa, A. Pentland (1997) “Facial expression recognition using image motion” in “Motion Based Recognition” M. Shah, R. Jain (eds) Dordrecht: Kluwer Acad. Pub. pp. 271–298CrossRefGoogle Scholar
  6. 6.
    A. Gianchetti, M. Campani, V. Torre (1998) “The use of optical flow for road navigation” IEEE trans Rob.Aut. 14 (1) pp 34–48CrossRefGoogle Scholar
  7. 7.
    E. Grosso, G. Metta, A. Oddera, G. Sandini (1996) “Robust visual sevoing in 3D reaching tasks” IEEE trans. Rob. Aut. 12 (5) pp. 732–741CrossRefGoogle Scholar
  8. 8.
    BKP Horn “Robot Vision” MIT Press (1986) (1990)Google Scholar
  9. 9.
    R.M. Haralick, L.G. Shapiro (1993) “Computer and robot vision” Reading, MA: Addison-WesleyGoogle Scholar
  10. 10.
    Jain A.K. (1989) “Fundamentals of digital image processing” Englewood-Clliffs: Prentice-HallGoogle Scholar
  11. 11.
    R. Jain, R. Kasturi, B.G. Schunck (1995) “ Machine Vision” New York: MacGraw HillGoogle Scholar
  12. 12.
    R. Klette, K. Schluns, A. Koshan (1999) “Computer vision. Three dimensional data from images” Berlin: Springer-VerlagGoogle Scholar
  13. 13.
    Kohonen T. (1995) “Self Organizing Maps” Berlin: Springer-VerlagGoogle Scholar
  14. 14.
    J. Konrad, E. Dubois (1992) “Bayesian estimation of motion vector fields” IEEE Patt. Anal. Mach. Int, 14 (9) pp. 910–927CrossRefGoogle Scholar
  15. 15.
    S. Maybank (1993) “Theory of reconstruction from image motion” Berlin: Springer VerlagGoogle Scholar
  16. 16.
    D. Nair, J.K. Aggarwal (1998) “Moving obstacle detection from a navigating robot” IEEE trans Rob. Aut. 14 (3) pp 404–416CrossRefGoogle Scholar
  17. 17.
    W.J. Niessen, J.S. Duncan, LMJ Florack, BM ter Haar Romeny, MA Viergever (1995) “Spatiotemporal operators and optic flow” Proc. Workshop on physics-based modelling in computer vision, IEEE Computer Soc. Press. pp 78–84Google Scholar
  18. 18.
    N. P. Papanikolopoulos, P.K. Koshla, T. Kanade (1993) “Visual tracking of a moving target by a camera mounted on a robot: A combination of control and vision” IEEE trans. Rob. Aut. 9 (1) pp. 14–34CrossRefGoogle Scholar
  19. 19.
    G.P. Stein, A, Shasua (2000) “Model based brightness constraints: on direct estimation of structure and motion” IEEE trans. Patt. Anal. Mach. Int. 22 (9) pp. 992CrossRefGoogle Scholar
  20. 20.
    C. Stiller, J. Konrad (1999) “Estimating motion in image sequences” IEEE Signal Processing Magazine, 16 (4),pp. 70–91CrossRefGoogle Scholar
  21. 21.
    M. Tistarelli (1996) “Multiple constraints to compute optical flow” IEEE Trans. Patt. Anal Mach. Int. 18 (12) pp. 1243–1250CrossRefGoogle Scholar
  22. 22.
    D. Vernon, M. Tistarelli (1990) “Using camera motion to estimate range for robotic parts manipulation” IEEE trans. Robot. Autom. 6 (5) pp. 509–521CrossRefGoogle Scholar
  23. 23.
    L. Wilxson (2000) “Detecting salient motion by accumulating directionally consistent flow” IEEE trans. Patt. Anal. Mach. Int. 22 (8) pp. 774CrossRefGoogle Scholar
  24. 24.
    Winkler G. (1995) “Image analysis, random fields and dynamic Monte Carlo methods” Berlin: Springer VerlagGoogle Scholar
  25. 25.
    Y. Yacoob, L.S. Davis (1997) “Estimating image motion using temporal multi-scale models of flow and acceleration” in “Motion Based Recognition” M. Shah, R. Jain (eds) Dordrecht: Kluwer Acad. Pub. pp. 17–37Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2003

Authors and Affiliations

  • Manuel Graña
    • 1
  1. 1.Dept. CCIAUPV/EHUSan SebastianSpain

Personalised recommendations