EEG based user recognition using BUMP modelling

摘要: In this paper the use of electroencephalogram (EEG) as biometric identifier is investigated. The EEG within framework has already been introduced in recent past although it not extensively analyzed. contribution we apply “bump” modelling analysis for feature extraction stage an identification framework, order to reduce huge amount data recorded through EEG. For purpose study rely on “resting state with eyes closed” protocol. employed database composed 36 healthy subjects whose signals have acquired ad hoc laboratory. Different electrodes configurations pertinent protocol considered. A classifier based Mahalanobis distance tested enrollment and their identification. An information fusion performed at score level shown improve correct classification performance. obtained results show that accuracy 99.69% can be achieved. It represents high degree accuracy, given current research biometrics.