1 June 2004 Iterative estimation of Dempster-Shafer's basic probability assignment: application to multisensor image segmentation
Author Affiliations +
Optical Engineering, 43(6), (2004). doi:10.1117/1.1737373
Abstract
Basic probability assignment (BPA) definition remains a difficult problem to apply Desmpter-Shafer evidence theory to practical applications such as in image processing. A new iterative approach of multisensor data fusion based on the Dempster-Shafer framework is proposed. BPAs, modeled by a Gaussian distribution, are estimated iteratively and in an unsupervised way using the fused image and the source images. Data fusion is performed at the pixel level. Results on synthetic and real images are presented to illustrate the effectiveness of the proposed fusion scheme. Limitations of the method are discussed.
Fabien Salzenstein, Abdel-Ouahab Boudraa, "Iterative estimation of Dempster-Shafer's basic probability assignment: application to multisensor image segmentation," Optical Engineering 43(6), (1 June 2004). http://dx.doi.org/10.1117/1.1737373
JOURNAL ARTICLE
7 PAGES


SHARE
Back to Top