Scientific topics : Uncertainty models and imperfect data processing.
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 details :
- Evolution of the Evidential C-Means algorithm (called MECM developped within Nasr Makni’s PhD thesis) 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).This work shows better results in prostat segmentation and tumor detection compared with 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.2010.202and Hedi Tabia’s PhD thesis). This work 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.
Pr. Olivier COLOT is member of the team "Signal & Image" of LAGIS.
Position : Full Professor.
Email : firstname.lastname@example.org
Phone : +33.(0)188.8.131.52.28.
Scientific activities : (...)