Fusion System Based on Belief Functions Theory and Approximated Belief Functions for Tree Species Recognition

Rihab Ben Ameur (1) , Lionel Valet (1) , Didier Coquin (1)

1 LISTIC - Laboratoire d'Informatique, Systèmes, Traitement de l'Information et de la Connaissance

In this paper, an information fusion system for tree species recognition through leaves is proposed. This approach consists in training sub-classifiers (Random forests) with attributes extracted from leaf photos. The database is incomplete, partial and some data is conflicting. A hierarchical fusion system based on Belief functions theory allows the fusion of data provided by different sub-classifiers. Different procedures for reducing computational complexity are tested.

Mots clés: Belief functions; sub-classification; conflict; imprecise information; belief function approximations

Domaines: Traitement des images [eess.IV]; Intelligence artificielle [cs.AI]; Vision par ordinateur et reconnaissance de formes [cs.CV]; Apprentissage [cs.LG]

Rihab Ben Ameur, Lionel Valet, Didier Coquin. Fusion System Based on Belief Functions Theory and Approximated Belief Functions for Tree Species Recognition . IEEE International Conference on Image Processing Theory, Tools and Applications, Dec 2016, Oulu, Finland. 6 p., ⟨10.1109/IPTA.2016.7820955⟩. ⟨hal-01415621⟩