A new approach to computation of curvature scale space image for shape similarity retrieval

  • Farzin Mokhtarian
  • Sadegh Abbasi
  • Josef Kittler
Poster Session C: Compression, Hardware & Software, Databases, Neural Networks, Object Recognition & Reconstruction
Part of the Lecture Notes in Computer Science book series (LNCS, volume 1311)


The Curvature Scale Space image is a multi-scale organisation of the inflection points of a closed planar curve as it is smoothed. It consists of several arch shape contours, each related to a concavity or a convexity of the curve. In our recent works, we have used the maxima of these contours as shape descriptors to find similar shapes in large image databases. In this paper, we address the problem of shallow concavities. They may give rise to large contours in the CSS image. These contours may then match contours corresponding to deep and wide concavities during the matching process.

We have introduced a method to adjust the height of every CSS image contour which will be less than the height of the actual contour in the CSS image. A new global parameter for closed curves is also introduced. This parameter, like the CSS image representation, is robust with respect to noise, scale and change in orientation.


Actual Contour Straight Line Segment Closed Curf Global Parameter Planar Curf 
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.
    Mokhtarian, F. and Mackworth, A. K. “A theory of multiscale, curvature-based shape representation for planar curves” in PAMI 1992, VOL. 14, pp789–805.Google Scholar
  2. 2.
    Mokhtarian, F. “Silhouette-based isolated object recognition through curvature scale space” in IEEE Trans. PAMI 1995, Vol.17, pp.539–544Google Scholar
  3. 3.
    Mokhtarian, F. Abbasi S. and Kittler J. “Robust and Efficient Shape Indexing through Curvature Scale Space” in Proc. BMVC'96. Edinburgh, September 1996, pp 53–62.Google Scholar
  4. 4.
    Mokhtarian, F. Abbasi S. and Kittler J. “Efficient and Robust Retrieval by Shape Content through Curvature Scale Space” in Proc. of the First International Workshop on Image Databases and Multimedia Search, Amsterdam, Aug 1996, pp 35–42.Google Scholar
  5. 5.
    Abbasi S., Mokhtarian, F. and Kittler J. “Reliable Classification of Chrysanthemum Leaves through Curvature Scale Space” in the proc. of the Scale-Space '97 Conference, July 1997, Utrecht, Netherlands.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1997

Authors and Affiliations

  • Farzin Mokhtarian
    • 1
  • Sadegh Abbasi
    • 1
  • Josef Kittler
    • 1
  1. 1.Centre for Vision Speech and Signal Processing Department of Electronic & Electrical EngineeringUniversity of SurreyGuildfordEngland

Personalised recommendations