Efficient region segmentation through ‘creep-and-merge’

  • Antranig Basman
  • Joan Lasenby
  • Roberto Cipolla
Poster Session A: Color & Texture, Enhancement, Image Analysis & Pattern Recognition, Segmentation
Part of the Lecture Notes in Computer Science book series (LNCS, volume 1310)


We present a novel architecture for region-based segmentation of stationary and quasi-stationary statistics, which is designed to function correctly under the widest range of conditions. It is robust to the extremes of region topology and connectivity, and automatically maintains region boundaries sampled to the minimum scale at which the region configuration can be determined with statistical confidence. The algorithm is deterministic, and when operating on images from within its domain of validity, contains no adjustable parameters. In contrast to most other techniques directed at the same problem, the progress of the algorithm cannot be described by the optimisation of a global energy criterion.

We describe a specific implementation using Gaussian stationary statistics, and present test results which demonstrate superior performance to a collection of other systems.


  1. 1.
    P. Andrey and P. Tarroux. ‘Unsupervised Texture Segmentation using Selectionist Relaxation'. In Proc. 4th European Conf. on Computer Vision, Volume I, pp. 482–491. Springer-Verlag, 1996. LNCS 1064.Google Scholar
  2. 2.
    H.J. Autonisse. Image segmentation in pyramids. Computer Vision, Graphics and Image Processing, 19(4):367–383, 1982.Google Scholar
  3. 3.
    J.A. Bangham, R. Harvey, P.D. Ling, and R.V. Aldridge. Nonlinear scale-space from n-dimensional sieves. In Proc. 4th European Conf. on Computer Vision, Volume I, pp. 189–198. Springer-Verlag, 1996. LNCS 1064.Google Scholar
  4. 4.
    A.M. Basman, J. Lasenby, and R. Cipolla. The Creep-and-Merge segmentation system. Technical Report CUED/F-INFENG/TR295, University of Cambridge, July 1997.Google Scholar
  5. 5.
    L.D. Cohen. On active contour models and balloons. Computer Vision, Graphics and Image Processing, 53(2):211–218, May 1991.Google Scholar
  6. 6.
    J.H. Elder and S.W. Zucker. Local scale control for edge detection and blur estimation. In Proc. 4th European Conf. on Computer Vision, Volume II, pp. 58–69. Springer-Verlag, 1996. Lecture Notes in Computer Science 1065.Google Scholar
  7. 7.
    S.A. Hojjatoleslami and J. Kittler. Region growing: A new approach. Technical report, University of Surrey, 1995.Google Scholar
  8. 8.
    H. Jiang, J. Toriwaki, and H. Suzuki. Comparative performance evaluation of segmentation methods based on region growing and division. Systems and Computers in Japan, 24(13):28–42, 1993.Google Scholar
  9. 9.
    M. Kass, A.P. Witkin, and D. Terzopoulos. Snakes: Active contour models. Int. Journal of Computer Vision, 1(4):321–331, Jan 1998.Google Scholar
  10. 10.
    K.V. Mardia, J.T. Kent, and J.M. Bibby. Multivariate Analysis. Academic Press, 1979.Google Scholar
  11. 11.
    T. Pavlidis. Structural Pattern Recognition, Chapter 5. Springer-Verlag, 1977.Google Scholar
  12. 12.
    P. Shroeter and J. Bigun. Hierarchical image segmentation by multi-dimensional clustering and orientation-adaptive boundary refinement. Pattern Recognition, 28(5):295–709, May 1995.Google Scholar
  13. 13.
    S.D. Silvey. Statistical Inference, Chapter 6. Penguin, 1970.Google Scholar
  14. 14.
    F. van der Heijden. Image Based Measurement Systems, Chapter 6. John Wiley & Sons, 1994.Google Scholar
  15. 15.
    S.C. Zhu and A. Yuille. ‘Region Competition: Unifying Snakes, Region Growing, and Bayes/MDL for: Multiband Image Segmentation'. IEEE Trans. Pattern Analysis and Machine Intell., 18(9):884–900, September 1996.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1997

Authors and Affiliations

  • Antranig Basman
    • 1
  • Joan Lasenby
    • 1
  • Roberto Cipolla
    • 1
  1. 1.Department of EngineeringUniversity of CambridgeCambridgeEngland

Personalised recommendations