Investigation of fuzzy neural networks for object recognition in electrooptical images
Abstract
The problem of object recognition in electrooptical images obtained using multispectral optical systems is considered. A fuzzy neural network is suggested for the solution of this problem. Different training algorithms for fuzzy neural network are considered, and the results of experimental investigations of their efficiency on the basis of real data are presented.Downloads
Published
2009-09-25
Issue
Section
Decision making and control in economic, technical, ecological and social systems