Drowsiness Detection Based on Eye Closure and Yawning Detection

Abstract

The number of major road accidents that occur per day is on a rise and most of them are attributed to being the driver’s fault. According to the survey done in 2015, drivers are held responsible for approximately 78% of the accidents. To minimize the occurrence of these incidents a monitoring system that alerts the driver when he succumbs to sleep is proposed. This algorithm processes live video feed focused on the driver’s face and tracks his eye and mouth movements to detect eye closure and yawning rates. An alarm sounds if the driver is drowsy or already asleep. Haar-cascade classifiers run parallelly on the extracted facial features to detect eye closure and yawning.