Fast and Exact Primal-Dual Iterations for Variational Problems in Computer Vision

  • Jan Lellmann
  • Dirk Breitenreicher
  • Christoph Schnörr
Part of the Lecture Notes in Computer Science book series (LNCS, volume 6312)


The saddle point framework provides a convenient way to formulate many convex variational problems that occur in computer vision. The framework unifies a broad range of data and regularization terms, and is particularly suited for nonsmooth problems such as Total Variation-based approaches to image labeling. However, for many interesting problems the constraint sets involved are difficult to handle numerically. State-of-the-art methods rely on using nested iterative projections, which induces both theoretical and practical convergence issues. We present a dual multiple-constraint Douglas-Rachford splitting approach that is globally convergent, avoids inner iterative loops, enforces the constraints exactly, and requires only basic operations that can be easily parallelized. The method outperforms existing methods by a factor of 4 − 20 while considerably increasing the numerical robustness.


Saddle Point Variational Problem Input Image Saddle Point Problem Segmentation Problem 
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.
    Alberti, G., Bouchitté, G., Dal Maso, D.: The calibration method for the Mumford-Shah functional and free-discontinuity problems. Calculus of Variations and Partial Differential Equations 16(16), 299–333 (2003)zbMATHCrossRefMathSciNetGoogle Scholar
  2. 2.
    Ambrosio, L., Fusco, N., Pallara, D.: Functions of Bounded Variation and Free Discontinuity Problems. Clarendon Press (2000)Google Scholar
  3. 3.
    Boykov, Y., Veksler, O., Zabih, R.: Fast approximate energy minimization via graph cuts. Patt. Anal. Mach. Intell. 23(11), 1222–1239 (2001)CrossRefGoogle Scholar
  4. 4.
    Chambolle, A., Cremers, D., Pock, T.: A convex approach for computing minimal partitions. Tech. Rep. 649, Ecole Polytechnique CMAP (2008)Google Scholar
  5. 5.
    Chan, T.F., Esedoḡlu, S., Nikolova, M.: Algorithms for finding global minimizers of image segmentation and denoising models. J. Appl. Math. 66(5), 1632–1648 (2006)zbMATHMathSciNetGoogle Scholar
  6. 6.
    Douglas, J., Rachford, H.H.: On the numerical solution of heat conduction problems in two and three space variables. Trans. of the AMS 82(2), 421–439 (1956)zbMATHMathSciNetGoogle Scholar
  7. 7.
    Eckstein, J.: Splitting Methods for Monotone Operators with Application to Parallel Optimization. Ph.D. thesis, MIT (1989)Google Scholar
  8. 8.
    Gabay, D.: Augmented Lagrangian Methods: Applications to the Numerical Solution of Boundary-Value Problems. In: Applications of the Method of Multipliers to Variational Inequalities, ch. IX, p. 299. North-Holland (1983)Google Scholar
  9. 9.
    Goldstein, T., Bresson, X., Osher, S.: Geometric applications of the split Bregman method: Segmentation and surface reconstruction. CAM Report 09-06, UCLA (2009)Google Scholar
  10. 10.
    Goldstein, T., Bresson, X., Osher, S.: Global minimization of Markov random fields with applications to optical flow. CAM Report 09-77, UCLA (2009)Google Scholar
  11. 11.
    Ishikawa, H.: Exact optimization for Markov random fields with convex priors. Patt. Anal. Mach. Intell. 25(10), 1333–1336 (2003)CrossRefGoogle Scholar
  12. 12.
    Lellmann, J., Becker, F., Schnörr, C.: Convex optimization for multi-class image labeling with a novel family of total variation based regularizers. In: Int. Conf. Comp. Vis (2009)Google Scholar
  13. 13.
    Lellmann, J., Kappes, J., Yuan, J., Becker, F., Schnörr, C.: Convex multi-class image labeling by simplex-constrained Total Variation. In: Tai, X.-C., Mørken, K., Lysaker, M., Lie, K.-A. (eds.) SSVM 2009. LNCS, vol. 5567, pp. 150–162. Springer, Heidelberg (2009)Google Scholar
  14. 14.
    Lellmann, J., Schnörr, C.: Continuous multiclass labeling approaches and algorithms. Tech. rep., Univ. of Heidelberg (February 2010),
  15. 15.
    Mumford, D., Shah, J.: Optimal approximations by piecewise smooth functions and associated variational problems. Comm. Pure Appl. Math. 42, 577–685 (1989)zbMATHCrossRefMathSciNetGoogle Scholar
  16. 16.
    Paragios, N., Chen, Y., Faugeras, O. (eds.): The Handbook of Mathematical Models in Computer Vision. Springer, Heidelberg (2006)Google Scholar
  17. 17.
    Pock, T., Cremers, D., Bischof, H., Chambolle, A.: An algorithm for minimizing the Mumford-Shah functional. In: Int. Conf. Comp. Vis. (2009)Google Scholar
  18. 18.
    Pock, T., Cremers, D., Bischof, H., Chambolle, A.: Global solutions of variational models with convex regularization. Tech. rep., Graz Univ. of Tech. (2009)Google Scholar
  19. 19.
    Rockafellar, R.: Convex Analysis. Princeton UP, Princeton (1970)zbMATHGoogle Scholar
  20. 20.
    Setzer, S.: Split Bregman algorithm, Douglas-Rachford splitting and frame shrinkage. In: Tai, X.-C., Mørken, K., Lysaker, M., Lie, K.-A. (eds.) SSVM 2009. LNCS, vol. 5567, pp. 464–476. Springer, Heidelberg (2009)Google Scholar
  21. 21.
    Strang, G.: Maximal flow through a domain. Math. Prog. 26, 123–143 (1983)zbMATHCrossRefMathSciNetGoogle Scholar
  22. 22.
    Zach, C., Gallup, D., Frahm, J.M., Niethammer, M.: Fast global labeling for real-time stereo using multiple plane sweeps. In: Vis. Mod. Vis. (2008)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2010

Authors and Affiliations

  • Jan Lellmann
    • 1
  • Dirk Breitenreicher
    • 1
  • Christoph Schnörr
    • 1
  1. 1.Image and Pattern Analysis Group & HCI, Dept. of Mathematics and Computer ScienceUniversity of Heidelberg 

Personalised recommendations