Maximum Membership Scale Selection
The use of multi-scale features is explored in the setting of supervised image segmentation by means of pixel classification. More specifically, we consider an interesting link between so-called scale selection and the maximum combination rule from pattern recognition. The parallel with scale selection is drawn further and a multi-scale segmentation method is introduced that relies on a per-scale classification followed by an over-scale fusion of these outcomes. A limited number of experiments is presented to provide some further understanding of the technique proposed.
KeywordsScale Space Combination Rule Quadratic Discriminant Analysis Scale Selection Maximum Membership
Unable to display preview. Download preview PDF.
- 2.Witkin, A.: Scale-space filtering: A new approach to multi-scale description. IEEE International Conference on Acoustics, Speech and Signal Processing 9 (1984)Google Scholar
- 11.Forstner, W., Gulch, E.: A fast operator for detection and precise location of distinct points, corners and centres of circular features. In: Proceedings of the ISPRS Intercommission Conference on Fast Processing of Photogrammetric Data, pp. 281–305 (1987)Google Scholar
- 13.de Bruijne, M.: Shape particle guided tissue classification. In: IEEE Workshop on Mathematical Methods in Biomedical Image Analysis (2006)Google Scholar
- 24.Olsen, O., Florack, L., Kuijper, A. (eds.): DSSCV 2005. LNCS, vol. 3753. Springer, Heidelberg (2005)Google Scholar