Driver Drowsiness Detection Using EEG Features

The objective of this paper is to discover the EEG (Electroencephalogram) features that expressed meaningful changes during drowsy driving state compared to the normal driving. For this purpose, 8 healthy male and female participants were recruited to conduct drowsy driving experiment in a fixed-base driving simulator, which reproduced the inside of the actual vehicle. The experimental scenario was driving a 37 km straight highway without any obstacles. The data obtained through this experiment were analyzed using brain wave analysis software. As a result, we found that the alpha RMS (Root mean square) and differentiated alpha RMS waves showed meaningful changes during drowsiness state compared to normal state. In addition, we suggested new brain activity index, which was composed of four brain waves that are alpha, beta, theta and delta, to amplify meaningful change in transition from normal state to drowsiness. The statistical significances of the selected EEG features were tested using One-way ANOVA (Analysis of variance). The result indicated that all three EEG features showed statistical significance (p < 0.005). In conclusion, this paper suggested EEG features which have high accuracy for drowsiness detection. Currently, EEG measurement equipment such as dry type and non-contact type is actively developed. Therefore, it is expected that the drowsiness prevention system using the EEG features will be available in the near future.