Title :
On the Effectiveness of EEG Signals as a Source of Biometric Information
Author :
Yang, Su ; Deravi, Farzin
Author_Institution :
Sch. of Eng. & Digital Arts, Univ. of Kent, Canterbury, UK
Abstract :
This paper presents a biometric person recognition system using electroencephalogram (EEG) signals as the source of identity information. Wavelet transform is used for extracting features from raw EEG signals which are then classified using a support vector machine and a knearestneighbour classifier to recognize the individuals. A number of stimuli are explored using up to 18 subjects to generate person-specific EEG patterns to explore which type of stimulus may achieve better recognition rates. A comparison between two kinds of tasks - motor movement and motor imagery - appears to indicate that imagery tasks show better and more stable performance than movement tasks. The paper also reports on the impact of the number and positioning of the electrodes on performance.
Keywords :
biomedical electrodes; biometrics (access control); electroencephalography; feature extraction; medical signal processing; signal classification; support vector machines; wavelet transforms; EEG signals; biometric information source; biometric person recognition system; electrode positioning; electroencephalogram signals; feature extraction; motor imagery task; motor movement task; person-specific EEG pattern generation; signal classification; stable performance; support vector machines; wavelet transforms; Accuracy; Electrodes; Electroencephalography; Feature extraction; Performance evaluation; Support vector machines; Testing;
Conference_Titel :
Emerging Security Technologies (EST), 2012 Third International Conference on
Conference_Location :
Lisbon
Print_ISBN :
978-1-4673-2448-9
