Biometric Authentication of Individuals using Electroencephalography (EEG)

Abstract

In this research we proposed a novel authentication method based on electroencephalogram (EEG) responses of individuals while writing their own signature and other´s signature. Two healthy male subjects (average years 27) participated in the experiments. In one task of the experiments, we asked them to sign their personal signatures on a paper within 2 seconds just after hearing a beep sound. This was repeated 24 times with an interval of 5 seconds. In addition, the same experiment was done but subjects were asked to sign a new unfamiliar signature, that is, other person´s signature. During these two tasks, EEG was recorded at 14 locations by using Emotiv EPOC neuro-headset. Comparison of the averages of the frequency powers after fast Fourier transformation of EEG data revealed the similarities and differences at different frequencies between two tasks. Average comparison was done for each individual channel in 10 frequency ranges within 4 to 43.5 Hz showed the significant differences
between the two groups are found in beta rhythm, particularly from 16 to 23.5 Hz at O1, O2, and P7 (p< 0.05; Two-way repeated ANOVA).