| | |
Summary: Regionalized random germs by a classification
for probabilistic watershed. Application:
angiogenesis imaging segmentation
Guillaume Noyel, Jes´us Angulo, and Dominique Jeulin
MINES ParisTech, CMM - Centre de Morphologie Math´ematique, Math´ematiques
et Syst`emes, 35 rue Saint Honor´e - 77305 Fontainebleau cedex, France
{noyel,angulo,jeulin}@cmm.ensmp.fr
Summary. New methods are presented to generate random germs regionalized by
a previous classification in order to use probabilistic watershed on hyperspectral
images. These germs are much more efficient than the standard uniform random
germs.
1 Introduction
Probabilistic watershed was introduced by Angulo and Jeulin [1] to detect
the contours of the widest and the most contrasted regions in images. The
obtained contours are more regular and significant than these associated to
the deterministic watershed. Probabilistic watershed was then extended to
hyperspectral images by Noyel, Angulo and Jeulin [5].
The standard stochastic WS consists in starting from uniform random
points germs as sources to flood the norm of a gradient in order to obtain
the associated contours. After repeating the process a large number of times,
|
