Title :
A Dempster-Shafer evidence theory-based approach to object classification on multispectral/hyperspectral images
Author :
Popov, M.A. ; Topolnitskiy, Maxim V.
Author_Institution :
Sci. Centre for Aerosp. Res. of the Earth, Kiev, Ukraine
Abstract :
The algorithm for object classification on multispectral/hyperspectral images based on the Dempster-Shafer evidence theory is represented. The algorithm allows detecting not only separate classes but also their composition, i.e. takes into account the “mixed” pixels inherent in the presence of medium spatial resolution images.
Keywords :
hyperspectral imaging; image classification; image resolution; inference mechanisms; Dempster-Shafer evidence theory; hyperspectral image; medium spatial resolution images; multispectral image; object classification; Accuracy; Bayes methods; Classification algorithms; Covariance matrices; Hyperspectral imaging; Dempster-Shafer evidence theory; multispectral / hyperspectral images; object classification;
Conference_Titel :
Digital Technologies (DT), 2014 10th International Conference on
Conference_Location :
Zilina
DOI :
10.1109/DT.2014.6868729
