Combing a Porcupine via Stereographic Direction Difusion
This paper addresses the problem of feature enhancement in noisy images when the feature is known to be constrained to a manifold. As an example, we study the problem of direction denoising. This problem was treated recently and several solutions were proposed. The various solutions share the same structure. They are composed of two terms: A difusion term and a projection term. Analytically, the solutions differ in the difusion part. The projection part is equivalent in all works. Yet, as it is often the case, the analytically equivalent projection terms differ from a numerical viewpoint. We present in this work a new parameterization of the problem that enables us to work always in a numerically stable way.
KeywordsNoisy Image Switching Point Riemannian Structure South Hemisphere Polyakov Action
Unable to display preview. Download preview PDF.
- 1.T. Chan and J. Shen, “Variational restoration of non-.at image features: Models and algorithms”, SIAM J. Appl. Math., to appear.Google Scholar
- 3.R. Kimmel and N. Sochen, “Orientation Difusion or How to Comb a Porcupine?”, special issue on PDEs in Image Processing, Computer Vision and Computer Graphics, Journal of Visual Communication and Image Representation, to appear.Google Scholar
- 5.E. Kreyszig, “Differential Geometry”, Dover Publications, Inc., New York, 1991.Google Scholar
- 8.N. Sochen and R.M. Haralick and Y.Y. Zeevi, “A Geometric functional for Derivatives Approximation” EE-Technion Report, April 1999.Google Scholar
- 9.N. Sochen and R. Kimmel and A.M. Bruckstein, “Difusions and confusions in signal and image processing”, accepted to Journal of Mathematical Imaging and Vision.Google Scholar
- 10.N. Sochen and R. Kimmel and R. Malladi, “From high energy physics to low level vision”, Report, LBNL, UC Berkeley, LBNL 39243, August, Presented in ONR workshop, UCLA, Sept. 5 1996.Google Scholar
- 12.N. Sochen, “Stochastic processes in vision I: From Langevin to Beltrami”, CCIT Technion technical report No. 245Google Scholar
- 13.N. Sochen and Y.Y. Zeevi, “Representation of colored images by manifolds embedded in higher dimensional non-Euclidean space”, Proc. IEEE ICIP’98, Chicago, 1998.Google Scholar
- 14.B. Tang and G. Sapiro and V. Caselles, “Direction difusion”, International Conference on Computer Vision, 1999.Google Scholar
- 15.B. Tang and G. Sapiro and V. Caselles, “Color image enhancement via chromaticity difusion” Technical report, ECE-University of Minnesota, 1999.Google Scholar