Classification crédale multi-cibles - Multi-target evidential classification

Samir Hachour (1) , François Delmotte (1) , Eric Lefevre (1) , David Mercier (1) , John Klein (2) , Jean-Marc Vannobel (2)

1 LGI2A - Laboratoire de Génie Informatique et d'Automatique de l'Artois
2 LAGIS-SI

In this paper, we study the problem of joint tracking and classification of several targets. Targets are considered to be known and sufficiently separated so that they cannot be confused. Our goal is to shed light on the contribution of credal classification over the Bayesian classification. Our job is to track the movement of targets using conventional Kalman filters and IMMs, and compare the performances of credal and Bayesian classification.

Domaines: Analyse de données, Statistiques et Probabilités [physics.data-an]

Samir Hachour, François Delmotte, Eric Lefevre, David Mercier, John Klein, et al.. Classification crédale multi-cibles - Multi-target evidential classification. Rencontres Francophones sur la Logique Floue et ses Applications, 2012, Compiègne, France. pp.201-208. ⟨hal-00767682⟩