Residual Images and Optical Flow Results

  • Andreas Wedel
  • Daniel Cremers


In real world motion estimation a major source of degradation in the estimated correspondence field comes about through illumination effects. Two obvious solutions to this problem are to either explicitly model the physical effect of illumination or to pre-filter the images so as to only preserve the illumination-independent information. The first approach is quite sophisticated because it involves the generation of a physical model of the observed scene, including the reconstruction of geometry, illumination, material properties and explicitly modeling the light transport. For realistic outdoor scenes this is currently infeasible. In this chapter we therefore revert to the second approach and propose to use the concept of residuals, which is the difference between an image and a smoothed version of itself, also known as structure–texture decomposition. Experimental results confirm that using residual input images for optical flow improves the accuracy of flow field estimates for image sequences that exhibit illumination effects.


Optical Flow Fundamental Matrix Illumination Change Textural Part Optical Flow Field 
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. 3.
    Aujol, J.F., Gilboa, G., Chan, T.F., Osher, S.J.: Structure-texture image decomposition: modeling, algorithms, and parameter selection. Int. J. Comput. Vis. 67(1) (2006). doi: 10.1007/s11263-006-4331-z
  2. 6.
    Baker, S., Roth, S., Scharstein, D., Black, M.J., Lewis, J.P., Szeliski, R.: A database and evaluation methodology for optical flow. In: Online-Proc. International Conference on Computer Vision, Rio de Janeiro, Brazil, October 2007 Google Scholar
  3. 13.
    Brox, T., Bruhn, A., Papenberg, N., Weickert, J.: High accuracy optical flow estimation based on a theory for warping. In: Proc. European Conference on Computer Vision, Prague, Czech Republic, pp. 25–36 (2004) Google Scholar
  4. 15.
    Brox, T., Bregler, C., Malik, J.: Large displacement optical flow. In: Proc. International Conference on Computer Vision and Pattern Recognition, Miami, FL, USA, June 2009 Google Scholar
  5. 32.
    Felsberg, M.: On the relation between anisotropic diffusion and iterated adaptive filtering. In: Pattern Recognition (Proc. DAGM), Munich, Germany, pp. 436–445 (2008) CrossRefGoogle Scholar
  6. 39.
    Haussecker, H., Fleet, D.: Estimating optical flow with physical models of brightness variation. IEEE Trans. Pattern Anal. Mach. Intell. 23(6), 661–673 (2001) CrossRefGoogle Scholar
  7. 61.
    Meyer, Y.: Oscillating patterns in image processing and in some nonlinear evolution equations. In: The Fifteenth Dean Jacquelines B. Lewis Memorial Lectures, March 2001 Google Scholar
  8. 62.
    Mileva, Y., Bruhn, A., Weickert, J.: Illumination-robust variational optical flow with photometric invariants. In: Pattern Recognition (Proc. DAGM), pp. 152–162 (2007) CrossRefGoogle Scholar
  9. 78.
    Rudin, L.I., Osher, S., Fatemi, E.: Nonlinear total variation based noise removal algorithms. Physica D 60(1–4), 259–268 (1992) MATHCrossRefGoogle Scholar
  10. 85.
    Stein, F.: Efficient computation of optical flow using the Census transform. In: Pattern Recognition (Proc. DAGM), Tübingen, Germany, pp. 79–86 (2004) CrossRefGoogle Scholar
  11. 90.
    Sun, D., Roth, S., Lewis, J.P., Black, M.J.: Learning optical flow. In: Proc. European Conference on Computer Vision, Marseille, France, pp. 83–91 (2008) Google Scholar
  12. 93.
    Torr, P.H.S., Zisserman, A.: Concerning Bayesian motion segmentation, model averaging, matching and the trifocal tensor. In: Proc. European Conference on Computer Vision, Freiburg, Germany, pp. 511–528 (1998) Google Scholar
  13. 95.
    Trobin, W., Pock, T., Cremers, D., Bischof, H.: An unbiased second-order prior for high-accuracy motion estimation. In: Pattern Recognition (Proc. DAGM), Munich, Germany, pp. 396–405 (2008) CrossRefGoogle Scholar
  14. 98.
    van de Weijer, J., Gevers, T.: Robust optical flow from photometric invariants. In: Int. Conf. on Image Processing, Singapore, pp. 1835–1838 (2004) Google Scholar
  15. 99.
    Vaudrey, T., Gruber, D., Wedel, A., Klappstein, J.: Space-time multi-resolution banded graph-cut for fast segmentation. In: Pattern Recognition (Proc. DAGM), Munich, Germany, pp. 203–213 (2008) CrossRefGoogle Scholar
  16. 104.
    Vidal, R., Sastry, S.: Optimal segmentation of dynamic scenes from two perspective views. In: Proc. International Conference on Computer Vision and Pattern Recognition, Madison, WI, USA, vol. 2, pp. 281–285 (2003) Google Scholar
  17. 107.
    Wedel, A., Pock, T., Zach, C., Cremers, D., Bischof, H.: An improved algorithm for TV-L1 optical flow. In: Workshop on Statistical and Geometrical Approaches to Visual Motion Analysis, Schloss Dagstuhl, Germany, September 2008 Google Scholar

Copyright information

© Springer-Verlag London Limited 2011

Authors and Affiliations

  1. 1.Group ResearchDaimler AGSindelfingenGermany
  2. 2.Department of Computer ScienceTechnical University of MunichGarchingGermany

Personalised recommendations