Segmentation as a Riemannian drum problem

In this paper, the segmentation problem is formulated as a problem of segmenting a Riemannian manifold. The image domain is endowed with an anisotropic metric and its segmentation is obtained by thresholding the second eigenvector of the Laplace-Beltrami operator on the Riemannian manifold so defined. The formulation is an analytic analog of a recently proposed approach to segmentation based on graph theory. However, the analytic formulation has built-in invariance properties and permits more general metrics. The formulation may also be viewed as a generalization of the method of curve evolution which is based on isotropic metrics.

[1]  Jitendra Malik,et al.  Normalized cuts and image segmentation , 1997, Proceedings of IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[2]  Bojan Mohar,et al.  Laplace eigenvalues of graphs - a survey , 1992, Discret. Math..

[3]  Gene H. Golub,et al.  Matrix computations , 1983 .

[4]  R. Merris Laplacian matrices of graphs: a survey , 1994 .

[5]  J. Cheeger A lower bound for the smallest eigenvalue of the Laplacian , 1969 .

[6]  Jayant Shah,et al.  A common framework for curve evolution, segmentation and anisotropic diffusion , 1996, Proceedings CVPR IEEE Computer Society Conference on Computer Vision and Pattern Recognition.