Scale-Based Corner Extraction of a Contour Figure Using a Crystalline Flow
We propose a scale-based method for extracting corners from a given polygonal contour figure. A crystalline flow is introduced to represent geometric features in a scale-space. It is an extension of a usual curvature flow. A special class of polygonal contours is evolved based on the nonlocal curvature. The nonlocal curvature is determined for each facet by its length. In the crystalline flow, a given polygon remains polygonal through the evolving process. Different from a usual curvature flow, it is easy to track a facet in a given polygon through the evolution. This aspect helps us to extract a set of dominant corners. Experimental results show that our method extracts a set of dominant corner facets successfully from a given contour figure.
KeywordsTransition Number Initial Contour Base Scale Polygonal Curve Dominant Facet
Unable to display preview. Download preview PDF.
- 3.T. Lindeberg, Scale-space theory in computer vision, Kulwer Academic(1994), in chap.8 218Google Scholar
- 4.J.A. Sethian, Level Set Methods: Evolving interfaces in geometry, Fluid Mechanics, Computer Vision, and Material Science, Cambridge(1996)Google Scholar
- 7.B. M. ter Haar Romeny edt. Geometry Driven Diffusion in Computer Vision, Kluwer Academic Publishers(1994)Google Scholar
- 8.J. Taylor, Constructions and conjectures in crystalline nondifferential geometry, In Differential Geometry (eds. B. Lawson and K. Tanenblat), Proceedings of the Conference on Differential Geometry, Rio de Janeiro, Pitman Monographs in Pure and Applied Math. 52(1991), pp.321–336Google Scholar
- 10.M.-H. Giga and Y. Giga, Motion driven by nonlocal curvature in the plane, In Asymptotics in Nonlinear Diffusive Systems, (eds. Y. Nishiura et al.), Tohoku Math. Publ. 8(1998), pp. 75–83Google Scholar