Abstract
Keywords: sleepiness, drowsy driving, image processing, eyelid behaviour, driver monitor
This study focuses on a driver sleepiness detection system in order to prevent drivers from accidents caused by drowsy driving. We conducted experiments on drowsy driving using a driving simulator. The system consists of a camera, an illuminator, and an image processing ECU. The ECU output the eyelid opening ratio, then we extracted several eyelid behaviours. We derived the sleepiness estimate function from these eyelid behaviours. As a result of verifying the accuracy, the average of correlation among 10 cases was over 0.90, and there was robust for the individual difference and the intraindividual difference.