Estimating 3D Flow for Driver Assistance Applications

  • Jorge A. Sánchez
  • Reinhard Klette
  • Eduardo Destefanis
Part of the Lecture Notes in Computer Science book series (LNCS, volume 5414)


This paper proposes a technique for estimating 3D flow vectors, by combining a KLT tracker with subsequent scale-space analysis of tracked points. A tracked point defines a 2D vector, which is mapped into 3D space based on ratios of maxima of scale-space characteristics. The approach is tested for night-vision sequences as recorded (at Daimler AG, Germany) for driver assistance projects. Those image sequences (at 25Hz) are characterized by being slightly blurry and of low contrast.


Motion analysis motion vector fields 3D motion driver assistance 


  1. 1.
    Deriche, R.: Recursively implementing the Gaussian and its derivatives. In: Proc. 2nd Int. Conf. on Image Processing, pp. 263–267 (1992)Google Scholar
  2. 2.
    enpeda..Image Sequence Analysis Test Site,
  3. 3.
    Intel Open Source Computer Vision Library,
  4. 4.
    Lindeberg, T.: On scale selection for differential operators. In: ISRN KTH/NA/P–93/12–SE, pp. 857–866 (1993)Google Scholar
  5. 5.
    Lindeberg, T.: Scale-Space Theory in Computer Vision. Kluwer Academic Publishers, Norwell (1994)CrossRefzbMATHGoogle Scholar
  6. 6.
    Lindeberg, T.: Feature detection with automatic scale selection. Int. J. Computer Vision 30, 77–116 (1998)Google Scholar
  7. 7.
    Liu, Z., Klette, R.: Performance evaluation of stereo and motion analysis on rectified image sequences. Technical report, Computer Science Department, The University of Auckland (2007)Google Scholar
  8. 8.
    Lowe, D.G.: Object Recognition from Local Scale-Invariant Features. In: Proc. ICCV, pp. 1150–1157 (1999)Google Scholar
  9. 9.
    Lucas, B., Kanade, T.: An iterative image registration technique with an application to stereo vision. In: Proc. IJCAI, pp. 674–679 (1981)Google Scholar
  10. 10.
    Matas, J., Chum, O., Martin, U., Pajdla, T.: Robust wide baseline stereo from maximally stable extremal regions. In: Proc. British Machine Vision Conference, vol. 1, pp. 384–393 (2002)Google Scholar
  11. 11.
    Mikolajczyk, K.: Detection of local features invariant to affine transformations, PhD hesis, Institut National Polytechnique de Grenoble, France (2002)Google Scholar
  12. 12.
    Shi, J., Tomasi, C.: Good Features to Track. In: Proc. IEEE Conf. Computer Vision Pattern Recognition, pp. 674–679 (1994)Google Scholar
  13. 13.
    Tomasi, C., Kanade, T.: Shape and Motion from Image Streams under Orthography: a Factorization Method. International Journal of Computer Vision 9, 137–154 (1992)CrossRefGoogle Scholar
  14. 14.
    van Vliet, L.J., Young, I., Verbeek, P.: Recursive Gaussian derivative filters. In: Proc. Int. Conf. on Pattern Recognition, pp. 509–514 (1998)Google Scholar
  15. 15.
    Vedula, S., Baker, S., Rander, P., Collins, R., Kanade, T.: Three-Dimensional Scene Flow. IEEE Transactions on Pattern Analysis and Machine Intelligence, 475–480 (2005)Google Scholar
  16. 16.
    Wedel, A., Rabe, C., Vaudrey, T., Brox, T., Franke, U., Cremers, R.: Efficient Dense Scene Flow from Sparse or Dense Stereo Data, Technical report, Computer Science Department, The University of Auckland (2008)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2009

Authors and Affiliations

  • Jorge A. Sánchez
    • 1
  • Reinhard Klette
    • 2
  • Eduardo Destefanis
    • 1
  1. 1.Facultad Regional CórdobaUniversidad Tecnológica NacionalCordobaArgentina
  2. 2.The .enpeda.. ProjectThe University of AucklandAucklandNew Zealand

Personalised recommendations