FEATURE EXTRACTION OF ARTIFICIAL TONGUE DATA USING GRAM-SCHMIDT ORTHONORMALIZATION (MonPmOR9)
Author(s) :
Iasen Hristozov (Bulgarian Academy of Science, Bulgaria)
Pencheva Tania (Bulgarian Academy of Science, Bulgaria)
Selim Eskiizmirliler (University of Denis Diderot, Paris VII, France)
Abstract : In this paper we present a combined feature extraction approach for an electronic tongue. The use of wavelet decomposition technique for feature extraction, followed by orthonormalization, decreases the number of classifier inputs to the multiplication of number of classes and number of sensors. This approach leads to a higher computational efficiency. Two experiments are presented to demonstrate the procedure.
Menu