Partenaires

CNRS
Ministère de l'Enseignement Supérieur et de la Recherche
Ecole Centrale de Lille
Université de Lille 1
CRIStAL

« avril 2017 »
L M M J V S D
27 28 29 30 31 1 2
3 4 5 6 7 8 9
10 11 12 13 14 15 16
17 18 19 20 21 22 23
24 25 26 27 28 29 30


Rechercher

Sur ce site

Sur le Web du CNRS


Accueil du site > Pages perso > Olivier COLOT > Scientific topics

Scientific topics

Collected data are tainted with several kinds of imperfections : uncertainty, imprecision, incompleteness, unreliability or conflict. In order to be efficiently processed, such data need to be represented by uncertainty models that best appraise what is actually known about them. To that end, we develop models in the framework of the belief function theory and fuzzy logic.

Some highlights (2008-2014) :

  • Evolution of the Evidential C-Means algorithm [1] (called MECM) and its use for the segmentation of multiparametric MRI of the prostat in the aim to identify the different tissues and to detect possible tumors (see for instance http://dx.doi.org/10.1118/1.3651610 , http://dx.doi.org/10.1016/j.inffus.2012.04.002 and Nasr Makni’s PhD thesis refered in TEL with identifier tel-00818024).This work shows better results in prostat segmentation and tumor detection compared to methods of the state-of-the-art.
  • The use of belief functions for matching and indexing of 3D patterns (see for instance http://dx.doi.org/10.1109/TPAMI.2010.202 and Hedi Tabia’s PhD thesis refered in TEL with identifier tel-00818224) shows better results in indexing and matching compared to the state-of-the-art.
  • Detection of singular sources based on the analysis of the conflict amount in belief functions combination (see http://dx.doi.org/10.1016/j.ijar.2011.08.005). This work shows better results compared to main methods of the state-of-the-art.

[1] initially developed by Masson and Denoeux.