Detection of drowsiness with eyes open using EEG-based power spectrum analysis

The number of traffic accident by drowsiness has the highest percentage every year. Recently, drowsiness detection is actively researched. The most practical approach to drowsy driving detection is generally known as image processing techniques which estimate drowsiness using eye-blink pattern. However, image processing technique can´t detect drowsiness with eyes open. This paper proposes method of drowsiness detection with eyes open using EEG-based power spectrum analysis. Subjects are selected that based on ESS(Epworth Sleepiness Scale). In experiments, all electronic devices are turned off to reduce the artifacts and noiseless environment was created to cause drowsiness. After the EEG experiments is complete, drowsy periods are classified according to alpha power spectrum changes that induced by eyes closed in drowsy state. Although subject´s eyes are opened for a long time, the drowsiness patterns are detected. Consequently, detection of drowsiness with eyes open is possible by using EEG-based power spectrum analysis.