- Startpagina tijdschrift
- Volume 14 (1995)
- Number 1 - Stermat '94 - Sep. 1995
- Conditional watershed segmentation for grain boundary reconstruction
Weergave(s): 271 (1 ULiège)
Download(s): 81 (1 ULiège)
Conditional watershed segmentation for grain boundary reconstruction
Abstract
This paper presents a new approach of watershed extraction in binary images. After a brief reminder about segmentation tools in mathematical morphology, the method, based on the characterisation of contour’s points, is detailed. The data used are given by the calypter which is a morphological descriptor incorporating skeleton and contour informations. Then, the advantages of the method are highlighted; in particular, it is shown that the geometrical knowledge, given by the calypter, about the size and configuration of the catchment basins can bring some criteria for the segmentation decision.
Om dit artikel te citeren:
Over : Eric Pirard
MICA-ULg, Université de Liège, Avenue des Tilleuls 45, 4000 Liège, Belgium
Over : Annick Delanaye
MICA-ULg, Université de Liège, Avenue des Tilleuls 45, 4000 Liège, Belgium