Biometric approach based on physiological human signals

Abstract

This study presents an approach to use different biomedical signals in order to do biometric identification. The biomedical signals are captured using Arduino and Libelium platforms, what offers a low cost solution. The biomedical signals used are Electromyogram, Electrocardiogram and the Galvanic Skin Response. These signals are parametrized using well-known measures and Neural Networks are used as classifier in order to develop the user identification. The result of a success rate of 85.55% what is understood as a promising way to identify people by their biomedical signals.