Advertisement

Fast Difference Schemes for Edge Enhancing Beltrami Flow and Subjective Surfaces

  • Ravi Malladi
  • Igor Ravve
Conference paper
Part of the Mathematics and Visualization book series (MATHVISUAL)

Summary

Numerical integration of space-scale PDE is one of the most time consuming operation of image processing. The scale step is limited by conditional stability of explicit schemes. In this work, we introduce the unconditionally stable semi-implicit linearized difference scheme for the Beltrami flow. The Beltrami flow [14, 15] is one of the most effective denoising algorithms in image processing. For gray-level images, we show that the Beltrami flow equation can be arranged in a reaction-diffusion form. This reveals the edge-enhancing properties of the equation and suggests the application of additive operator split (AOS) methods [4, 5] for faster convergence. As we show with numerical simulations, the AOS method results in an unconditionally stable semi-implicit linearized difference scheme in 2D and 3D. The values of the edge indicator function are used from the previous step in scale, while the pixel values of the next step are used to approximate the flow. The optimum ratio between the reaction and diffusion counterparts of the governing PDE is studied. We then apply this approach to fast subjective surface computation. The computational time decreases by a factor of 20 and more, as compared to the explicit scheme.

Keywords

Beltrami flow subjective surfaces unconditionally stable scheme segmentation 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    P. Perona, J. Malik, “Scale-space and edge-detection using anisotropic diffusion,” IEEE Trans. of PAMI, Vol. 12, pp. 629–639, 1990.CrossRefGoogle Scholar
  2. 2.
    L. Alvarez, P.L. Lions, J.M. Morel, “Image selective smoothing and edge detection by non-linear diffusion II,” SIAM Journal on Numerical Analysis, Vol. 29 (3), pp. 845–866, 1992.MathSciNetzbMATHCrossRefGoogle Scholar
  3. 3.
    L. Rudin, S. Osher, and E. Fatemi, “Nonlinear total variation based noise removal algorithms,” Physica D, Vol. 60, pp. 259–268, 1992.zbMATHCrossRefGoogle Scholar
  4. 4.
    J. Weickert, B.M. ter Haar Romeny, and M.A. Viergever. “Efficient and reliable scheme for nonlinear diffusion filtering”. IEEE Trans on Image Processing., Vol. 7 (3), pp. 398–410 (1998).CrossRefGoogle Scholar
  5. 5.
    R. Goldenberg, R. Kimmel, E. Rivlin, and M. Rudzsky. “Fast Geodesic Active Contours”. M. Nielsen, P. Johansen, O.F. Olsen, J. Weickert (Editors), Scale-space theories in computer vision, Lecture Notes in Computer Science, Vol. 1682, Springer, Berlin, 1999.Google Scholar
  6. 6.
    S. Osher and J. A. Sethian, “Fronts propagating with curvature dependent speed: Algorithms based on Hamilton-Jacobi formulation,” Journal of Computational Physics, Vol. 79, pp. 12–49, 1988.MathSciNetzbMATHCrossRefGoogle Scholar
  7. 7.
    R. Malladi and I. Ravve, “Fast difference schemes for edge-enhancing Beltrami flow,” in Proceedings of ECCV, LNCS 2350, pp. 343–357, Copenhagen, Denmark, May 2002.Google Scholar
  8. 8.
    R. Malladi and J. A. Sethian, “Image processing via level set curvature flow,” Proc. of Natl. Acad. of Scie., Vol. 92, pp. 7046–7050, July 1995.MathSciNetzbMATHCrossRefGoogle Scholar
  9. 9.
    R. Malladi and J. A. Sethian, “Image Processing: Flows under Min/Max curvature and Mean curvature,” Graphical Models and Image Processing, Vol. 58 (2), pp. 127–141, 1996.MathSciNetCrossRefGoogle Scholar
  10. 10.
    G.I. Barenblatt. “Self-Similar Intermediate Asymptotics for Nonlinear Degenerate Parabolic Free-Boundary Problems which Occur in Image Processing”. to appear in Proceedings of the National Academy of Science, 2001.Google Scholar
  11. 11.
    B. M. ter Harr Romeny, editor, Geometry-driven diffusion in computer vision, Kluwer Academic publishers, The Netherlands, 1994.Google Scholar
  12. 12.
    R. Kimmel, R. Malladi, and N. Sochen. “Images as embedded maps and minimal surfaces: movies, color, texture, and volumetric medical images”. International Journal of Computer Vision, 1999.Google Scholar
  13. 13.
    R. Kimmel, “A natural norm for color processing,” in Proceedings of 3-rd ACCV, Springer-Verlag LNCS 1352, pp. 88–95, Hong-Kong, January 1998.Google Scholar
  14. 14.
    N. Soeben, R. Kimmel and R. Malladi, “From high energy physics to low level vision,” LBNL Report # 39243, Lawrence Berkeley National Laboratory, University of California, Berkeley, Aug. 1996.Google Scholar
  15. 15.
    N. Sochen, R. Kimmel and R. Malladi, “A general framework for low level vision”. IEEE Transactions, Vol. 7, No. 3, 1998.Google Scholar
  16. 16.
    G. Sapiro, “Vector-valued active contours,” in Proceedings of CVPR ‘86, pp. 650–655, 1996.Google Scholar
  17. 17.
    A. M. Polyakov, “Quantum geometry of bosonic strings,” in Physics Letters B, 103B(3), pp. 207–210, 1981.Google Scholar
  18. 18.
    R. Malladi and I. Ravve, “Fast difference scheme for anisotropic Beltrami smoothing and edge contrast enhancement of gray level and color images,” LBNL report # 48796, Lawrence Berkeley National Laboratory, University of California, Berkeley, August, 2001.Google Scholar
  19. 19.
    J. Weickert. Anisotropie Diffusion in Image Processing. Ph. D. Thesis, Kaiserslautern University, 1996.Google Scholar
  20. 20.
    Curtis E. Gerald and Patric O. Wheatley. Applied Numerical Analysis, Addison Wesley, NY, 1999.Google Scholar
  21. 21.
    William H. Press, Saul A. Tekolsky, William T. Vetterling and Brian F. Flannery. Numerical Recipes in C, Cambridge University Press, 1992Google Scholar
  22. 22.
    A. Yezzi, “Modified curvature motion for image smoothing and enhancement,” IEEE Tran. of Image Processing, Vol. 7, No. 3, 1998.Google Scholar
  23. 23.
    A. Sarti, R. Malladi, and J. A. Sethian, “Subjective surfaces: A method for completing missing boundaries,” Proc. Natl. Acad. Sci. USA, Vol. 97, No. 12, 2000.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2003

Authors and Affiliations

  • Ravi Malladi
    • 1
  • Igor Ravve
    • 1
  1. 1.Lawrence Berkeley National Laboratory, Computing Science DepartmentUniversity of CaliforniaBerkeleyUSA

Personalised recommendations