Target Tracking Using Interacting Multiple Models with Particle Filtering

ALBERTO M. CORRAL; BRUNO CERNUSCHI FRÍAS

Córdoba

Congreso; 12th Argentine Symposium on Technology, AST-2011, como parte de la 40 Conferencia Argentina en Informática, JAIIO 40, SADIO, 29 de Agosto al 2 de Septiembre de 2011.; 2011

Universidad Tecnológica Nacional, Regional Córdoba y Sociedad AArgentina de Informática, SADIO.

The problem of modeling accuracy in target tracking hasbeen well studied in the past and is specially important when trackingmaneuvering targets. One of the most simple and elegant ways ofimproving an algorithm in this sense is by using Interacting MultipleModel (IMM). IMM is a method that takes into account more than onemodel at the same time. This paper describes how it works and how ithas been incorporated in tracking algorithms in the past, specially in theExtended Kalman Filter (EKF). We also introduce a novel way of usingit with Particle Filters (PF). The original proposal found here is thatwe estimate the whole target state sampling particles from the OptimalFunction.