Automated video-based measurement of eye closure for detecting behavioral microsleep

A device capable of continuously monitoring an individual's levels of alertness in real-time is highly desirable for preventing drowsiness and microsleep related accidents. This paper presents a development of non-intrusive and light-insensitive video-based system that uses computer-vision methods to measure facial metric for identifying visible facial signs of drowsiness and behavioral microsleep. The developed system uses a remotely placed camera with a near-infrared illumination to acquire the video. The computer-vision methods are then applied to sequentially localize face, eyes, and eyelids positions to measure ratio of eye closure. The system was evaluated in frontal images of nine subjects with varying facial structures and exhibiting several ratio of eye closure and eye gaze under fully dark and ambient lighting conditions. The preliminary results showed promising results with sufficient accuracy to distinguish between fully closed, half closed, and fully open eyes.