Affine invariant deformation curves a tool for shape characterization

Abstract In this paper we describe a combination of two well-known concepts: morphological deformation curves and affine invariant evolution processes. A morphological deformation curve is a shape descriptor obtained by continuously deforming the shape and in the meantime measuring some geometric parameter. This measurement as a function of the deformation provides a characterization of the shape that can be used for shape classification and recognition. In classical morphology the deformation process is taken to be a morphological operator (erosions, dilations and combinations thereof). Most of the morphological operators are only invariant under the Euclidean symmetry group (translation and rotation) and (relative) invariant under isotropic rescalings. Only recently in the context of computer vision, an affine invariant shape deformation process has been identified. When we measure the area of the object while it is deformed in this way we obtain a function that provides a characterization of a shape in an affine invariant way. In this paper an affine invariant deformation curve is introduced and its application for shape characterization is illustrated.

[1]  Petros Maragos,et al.  Pattern Spectrum and Multiscale Shape Representation , 1989, IEEE Trans. Pattern Anal. Mach. Intell..

[2]  Leo Dorst,et al.  Quadratic Structuring Functions in Mathematical Morphology , 1996, ISMM.

[3]  Joachim Weickert,et al.  Anisotropic diffusion in image processing , 1996 .

[4]  Arnold W. M. Smeulders,et al.  The morphological structure of images , 1992, Proceedings., 11th IAPR International Conference on Pattern Recognition. Vol. III. Conference C: Image, Speech and Signal Analysis,.

[5]  Jean Serra,et al.  Image Analysis and Mathematical Morphology , 1983 .

[6]  Rangachar Kasturi,et al.  Machine vision , 1995 .

[7]  G. Matheron.  Random Sets and Integral Geometry , 1976 .

[8]  Jean-Michel Morel,et al.  Morphological Approach to Multiscale Analysis: From Principles to Equations , 1994, Geometry-Driven Diffusion in Computer Vision.

[9]  Edward R. Dougherty,et al.  An introduction to morphological image processing , 1992 .

[10]  Max A. Viergever,et al.  Numerical Analysis of Geometry-Driven Diffusion Equations , 1994, Geometry-Driven Diffusion in Computer Vision.

[11]  S. Zucker,et al.  Toward a computational theory of shape: an overview , 1990, eccv 1990.