Multisource classification using ICM and Dempster-Shafer theory

Author

Foucher, Samuel ; Germain, Mickaël ; Boucher, Jean-Marc ; Bénié, Goze Bertin

Author_Institution

Centre d´´Applications et de Recherche en Teledetection, Sherbrooke Univ., Que., Canada

Volume

51

Issue

2

fYear

2002

fDate

4/1/2002 12:00:00 AM

Firstpage

277

Lastpage

281

Abstract

We propose to use evidential reasoning in order to relax Bayesian decisions given by a Markovian classification algorithm, the multiscale iterated conditional mode (ICM) algorithm. The Dempster-Shafer rule of combination enables us to fuse decisions in a local spatial neighborhood which we further extend to be multisource. This approach enables us to more directly fuse information. Application to the classification of very noisy images produces interesting results

Keywords

Bayes methods; Markov processes; image classification; iterative methods; radar imaging; sensor fusion; uncertainty handling; Bayesian decisions; Dempster-Shafer combination rule; Dempster-Shafer theory; ICM; Markovian classification algorithm; data fusion; decision fusion; evidential reasoning; multiscale iterated conditional mode algorithm; multisource classification; multisource local spatial neighborhood; noisy image classification; radar image classification; remote sensing; Bayesian methods; Classification algorithms; Fuses; Image processing; Laser radar; Mathematical model; Optical sensors; Pixel; Radar imaging; Remote sensing;

fLanguage

English

Journal_Title

Instrumentation and Measurement, IEEE Transactions on

Publisher

ieee

ISSN

0018-9456

Type

jour

DOI

10.1109/19.997824

Filename

997824