Summary: Contour and Detail Detection for Spatially Adaptive Color Me-
F. Robert-Inacio1, J. Angulo2 and ´E. Dinet3
1Institut Mat´eriaux Micro´electronique et Nanosciences de Provence UMR CNRS 6242
Institut Sup´erieur de l' ´Electronique et du Num´erique, Toulon, France.
2Centre de Morphologie Math´ematique, Ecole des Mines de Paris, Fontainebleau, France
3Universit´e Jean Monnet, Laboratoire d'Informatique Graphique et d'Ing´enierie de la Vision, Saint- ´Etienne, France.
In this paper, we present a spatially adaptive filter for color
median filtering. Several alternative implementations of such a
filter are given, using different ways of gradient computation.
Then, several methods of color median filtering are compared
in terms of performance on a particular image. Vector median
filters are considered as well as median filters inspired by grey-
Median filtering has been widely explored since the last ten
years and many methods have been presented in order to filter
color images. Some of them are based on vector considerations
 and we will compare them to ours. First of all, we recall the