Monotonic additive preference model for 3D fusion system parameters adjustment
Résumé
3D image interpretation to understand complex phenomenon is achieved thanks to fusion systems having numerous parameters, difficult to adjust. An approximate model is looking for to simulate the 3D fusion process. The problem is described as a ranking problem and three MCDA methods are considered thanks to holistic preference information on a set of reference pictures: The ACUTA method with linear utilities, the ACUTA enriched by the consideration of linearity pieces and the UTA GMS method. Obtained results show the limit of using monotonic additive utilities for such identification problem.