Advertisement

Markovian Energy-Based Computer Vision Algorithms on Graphics Hardware

  • Pierre-Marc Jodoin
  • Max Mignotte
  • Jean-François St-Amour
Part of the Lecture Notes in Computer Science book series (LNCS, volume 3617)

Abstract

This paper shows how Markovian segmentation algorithms used to solve well known computer vision problems such as motion estimation, motion detection and stereovision can be significantly accelerated when implemented on programmable graphics hardware. More precisely, this contribution exposes how the parallel abilities of a standard Graphics Processing Unit (usually devoted to image synthesis) can be used to infer the labels of a label field. The computer vision problems addressed in this paper are solved in the maximum a posteriori (MAP) sense with an optimization algorithm such as ICM or simulated annealing. To do so, the fragment processor is used to update in parallel every labels of the segmentation map while rendering passes and graphics textures are used to simulate optimization iterations. Results show that impressive acceleration factors can be reached, especially when the size of the scene, the number of labels or the number of iterations is large. Hardware results have been obtained with programs running on a mid-end affordable graphics card.

Keywords

Simulated Annealing Motion Estimation Motion Detection Graphic Hardware Texture Memory 
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.

References

  1. 1.
    Moreland, K., Angel, E.: The fft on a gpu. In: Proc. of Workshop on Graphics Hardware, pp. 112–119 (2003)Google Scholar
  2. 2.
    Kruger, J., Westermann, R.: Linear algebra operators for gpu implementation of numerical algorithms. ACM Trans. Graph. 22(3), 908–916 (2003)CrossRefGoogle Scholar
  3. 3.
  4. 4.
    Strzodka, R., Rumpf, M.: Level set segmentation in graphics hardware. In: Proc. of ICIP, vol. 3, pp. 1103–1106 (2001)Google Scholar
  5. 5.
    Bouthemy, P., Lalande, P.: Motion detection in an image sequence using gibbs distributions. In: Proc. of ICASSP, pp. 1651–1654 (1989)Google Scholar
  6. 6.
    Konrad, J., Dubois, E.: Bayesian estimation of motion vector fields. IEEE Trans. Pattern Anal. Mach. Intell. 14(9), 910–927 (1992)CrossRefGoogle Scholar
  7. 7.
    Scharstein, D., Szeliski, R., Zabih, R.: A taxonomy and evaluation of dense two-frame stereo correspondence algorithms. In: Proc. of the IEEE Workshop on Stereo and Multi-Baseline Vision (2001)Google Scholar
  8. 8.
    Geman, S., Geman, D.: Stochastic relaxation, gibbs distributions, and the bayesian restoration of images. IEEE Trans. Pattern Anal. Machine Intell. 6(6), 721–741 (1984)zbMATHCrossRefGoogle Scholar
  9. 9.
    Kirkpatrick, S., Gelatt, C., Vecchi, M.: Optimization by simulated annealing. Science 220(4598), 671–680 (1983)CrossRefMathSciNetGoogle Scholar
  10. 10.
    Besag, J.: On the statistical analysis of dirty pictures. J. Roy. Stat. Soc. 48(3), 259–302 (1986)zbMATHMathSciNetGoogle Scholar
  11. 11.
    Chou, P., Brown, C.: The theory and practice of bayesian image labeling. In: Proc. of ICCV, pp. 185–210 (1990)Google Scholar
  12. 12.
    Dumontier, C., Luthon, F., Charras, J.-P.: Real-time dsp implementation for mfr-based video motion detection. IEEE Trans. on Img. Proc. 8(10), 1341–1347 (1999)CrossRefGoogle Scholar
  13. 13.
    Nagel, H.-H.: Image sequence evaluation: 30 years and still going strong. In: Proc. of ICPR, pp. 1149–1158 (2000)Google Scholar
  14. 14.
    Mitiche, A., Bouthemy, P.: Computation and analysis of image motion: a synopsis of current problems and methods. Int. J. Comput. Vision 19(1), 29–55 (1996)CrossRefGoogle Scholar
  15. 15.
    Black, M., Anandan, P.: The robust estimation of multiple motions: parametric and piecewise-smooth flow fields. Comput. Vis. Image Underst. 63(1), 75–104 (1996)CrossRefGoogle Scholar
  16. 16.
    Rost, R.J.: OpenGL Shading Language, 1st edn. Addison-Wesley, Reading (2004)Google Scholar
  17. 17.
    Akenine-Möller, T., Haines, E.: Real-time Rendering, vol. 2. AK Peters (2002)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2005

Authors and Affiliations

  • Pierre-Marc Jodoin
    • 1
  • Max Mignotte
    • 1
  • Jean-François St-Amour
    • 1
  1. 1.DIROUniversité de MontréalMontréal

Personalised recommendations