Curvature Dependent Skeletonization
The Hamilton-Jacobi approach has proved to be a powerful and elegant method for extracting the skeleton of a shape. The approach is based on the fact that the dynamics of the inward evolving boundary is conservative everywhere except at skeletal points. Nonetheless this method appears to overlook the fact that the linear density of the evolving boundary front is not constant where the front is curved. In this paper we present an analysis which takes into account variations of density due to boundary curvature. This yields a skeletonization algorithm that is both better localized and less susceptible to boundary noise than the Hamilton-Jacobi method.
KeywordsEikonal Equation Boundary Front Skeletonization Algorithm Picture Show Boundary Noise
- 3.D. Macrini et al. View-Based 3-D Object Recognition using Shock Graphs. In ICPR, pages 24–28. IEEE Computer Society, 2002.Google Scholar
- 6.M. Schmitt. Some examples of algorithms analysis in computational geometry by means of mathematical morphological techniques. In Geometry and Robotics, 1989. LNCS 391.Google Scholar
- 8.K. Siddiqi and B. B. Kimia. A shock grammar for recognition. In CVPR, pages 507–513. IEEE Computer Society, June 1996.Google Scholar
- 9.H. Tek and B. B. Kimia. Symmetry maps of free-form curve segments via wave propagation. In ICCV, pages 362–369. IEEE Computer Society, September 1999.Google Scholar