Exploring Non-linear Difusion: The Difusion Echo
The Gaussian serves as Green’s function for the linear diffusion equation and as a source for intuitive understanding of the linear difusion process. In general, non-linear difusion equations have no known closed formsolu tions and thereby no equally simple description. This article introduces a simple, intuitive description of these processes in terms of the Difusion Echo. The Difusion Echo offers intuitive visualisations for non-linear difusion processes.
In addition, the Difusion Echo has potential for o?ering simple formulations for grouping problems. Furthermore, the Difusion Echo can be considered a deep structure summary and thereby offers an alternative to multi-scale linking and ?ooding techniques.
KeywordsIntuitive Understanding Segmentation Task Watershed Segmentation Soft Threshold Intuitive Description
Unable to display preview. Download preview PDF.
- 1.The internet brain segmentation repository, 1999. MR brain data set 788⁃6⁃m and its manual segmentation was provided by the Center for Morphometric Analysis at MGH, http://neuro-www.mgh.harvard.edu/cma/ibsr.
- 3.Erik Dam. Evaluation of difusion schemes for watershed segmentation. Master#x2019;s thesis, University of Copenhagen, 2000. Technical report 2000/1 on http://www.diku.dk/research/techreports/2000.htm.
- 4.Erik Dam and Mads Nielsen. Non-linear difusion for interactive multi-scale watershed segmentation. MICCAI 2000, vol 1935 of LNCS, 216–225. Springer, 2000.Google Scholar
- 6.Tony Lindeberg. Scale-Space Theory in Computer Vision. Kluwer, 1994.Google Scholar
- 8.H. Scharr and J. Weickert. An anisotropic difusion algorithmwit h optimized rotation invariance. Mustererkennung 2000, DAGM, pp 460–467. Springer, 2000.Google Scholar
- 9.Joachim W eickert. Anisotropic Difusion in Image Processing. Teubner, 1998.Google Scholar
- 10.Andrew P. Witkin. Scale-space filtering. In Proceedings of International Joint Conference on Artificial Intelligence, pages 1019–1022, Karlsruhe, Germany, 1983.Google Scholar