ASSESSING SELF-SIMILARITY AND CROSS-SIMILARITY BETWEEN EEG PATTERNS FOR BIOMETRICAL APPLICATIONS
In the modern world, authentication and access control mechanisms have become a part of our daily lives. Although traditional methods are widely used, this report addresses the necessity of a more robust approach for biometric application in the access control technology.
Biometrics is the science of measuring and analyzing certain unique, physiological and behavioral human characteristics, called biometric identifiers, for authentication purposes. Specifically, this research investigates the advantage of using the Electroencephalogram (EEG) as a biometric identifier; as EEG may potentially improve the robustness and security, of biometric systems. EEG results from the electrical activity due to ionic current flows within the neurons of a functioning brain. The power spectral density estimates of EEG over the combined Alpha-Beta rhythm was used as the potential discriminant feature for biometric authentication of different subjects, each performing two different activities. To classify subjects, their recorded EEG were processed and analyzed through techniques, such as power spectral density (PSD) estimation and Euclidean distance classification. Although our approach successfully classified one subject for both predesigned mental tasks accurately, it has failed to show the sufficient accuracy for the other subject.