A Variational Framework for Joint Image Registration, Denoising and Edge Detection

  • Jingfeng Han
  • Benjamin Berkels
  • Martin Rumpf
  • Joachim Hornegger
  • Marc Droske
  • Michael Fried
  • Jasmin Scorzin
  • Carlo Schaller
Part of the Informatik aktuell book series (INFORMAT)

Abstract

In this paper we propose a new symmetrical framework that solves image denoising, edge detection and non-rigid image registration simultaneously. This framework is based on the Ambrosio-Tortorelli approximation of the Mumford-Shah model. The optimization of a global functional leads to decomposing the image into a piecewise-smooth representative, which is the denoised intensity function, and a phase field, which is the approximation of the edge-set. At the same time, the method seeks to register two images based on the segmentation results. The key idea is that the edge set of one image should be transformed to match the edge set of the other. The symmetric non-rigid transformations are estimated simultaneously in two directions. One consistency functional is designed to constrain each transformation to be the inverse of the other. The optimization process is guided by a generalized gradient flow to guarantee smooth relaxation. A multi-scale implementation scheme is applied to ensure the efficiency of the algorithm. We have performed preliminary medical evaluation on T1 and T2 MRI data, where the experiments show encouraging results.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Zöllei L, Yezzi A, Kapur T. A Variational Framework for Joint Segmentation and Registration. In: MMBIA’01: Proceedings of the IEEE Workshop on Mathematical Methods in Biomedical Image Analysis. Washington, DC, USA: IEEE Computer Society; 2001. p. 44–51.Google Scholar
  2. 2.
    Chen Y, Thiruvenkadam S, Huang F, Gopinath KS, Brigg RW. Simultaneous segmentation and registration for functional MR images. In: Proceedings. 16th International Conference on Pattern Recognition. vol. 1; 2002. p. 747–750.Google Scholar
  3. 3.
    Young Y, Levy D. Registration-based morphing of active contours for segmentation of CT scans. Mathematical Biosciences and Engineering 2005;2(1):79–96.MathSciNetGoogle Scholar
  4. 4.
    Pohl KM, Fisher J, Levitt JJ, Shenton ME, Kikinis R, Grimson WEL, et al. A Unifying Approach to Registration, Segmentation, and Intensity Correction. In: MICCAI; 2005. p. 310–318.Google Scholar
  5. 5.
    Droske M, Ring W. A Mumford-Shah Level-Set Approach for Geometric Image Registration. SIAM Appl Math 2005; to appear.Google Scholar
  6. 6.
    Droske M, Ring W, Rumpf M. Mumford-Shah Based Registration. Computing and Visualization in Science manuscript 2005; submitted.Google Scholar
  7. 7.
    Ambrosio L, Tortorelli VM. On the approximation of free discontinuity problems. Boll Un Mat Ital B 1992;6(7):105–123.MathSciNetGoogle Scholar
  8. 8.
    Ambrosio L, Tortorelli VM. Approximation of functionals depending on jumps by elliptic functionals via Γ-convergence. Comm Pure Appl Math 1990;43:999–1036.MathSciNetGoogle Scholar
  9. 9.
    Mumford D, Shah J. Boundary detection by minimizing functional. In: Proceedings. IEEE conference on Computer Vision and Pattern Recognition. San Francisco, USA; 1985.Google Scholar
  10. 10.
    Bourdin B, Chambolle A. Implementation of an adaptive Finite-Element approximation of the Mumford-Shah functional. Numer Math 2000;85(4):609–646.MathSciNetCrossRefGoogle Scholar
  11. 11.
    Clarenz U, Henn S, Rumpf K M Witsch. Relations between optimization and gradient flow methods with applications to image registration. In: Proceedings of the 18th GAMM Seminar Leipzig on Multigrid and Related Methods for Optimisation Problems; 2002. p. 11–30.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2006

Authors and Affiliations

  • Jingfeng Han
    • 1
  • Benjamin Berkels
    • 2
  • Martin Rumpf
    • 2
  • Joachim Hornegger
    • 1
  • Marc Droske
    • 3
  • Michael Fried
    • 1
  • Jasmin Scorzin
    • 2
  • Carlo Schaller
    • 2
  1. 1.Universität Erlangen-NürnbergDeutschland
  2. 2.Universität BonnBonn
  3. 3.University of CaliforniaUSA

Personalised recommendations