A Review on the Driver’s Fatigue Detection Methods

Abstract

Monitoring a driver’s fatigue is an essential factor in preventing accidents on the road. Several fatal accidents are caused by the driver’s sleep deprivation and exhaustion. Nowadays different systems exist that tell if the driver is getting drowsy. They are many signs that show on the driver’s facial features that indicate fatigue such as closed eyes, lower blinking frequency and yawning, but also a sudden change in the steering pattern, a lane deviation, a looser grip on the steering wheel, head tilting, leaning forward or sudden change of speed. Therefore, detecting these signs would help indicate the fatigue state of a driver. In this paper, we present a review of the different existing techniques for eye closure detection and yawning detection. We will discuss various algorithms used for face detection, eye and mouth detection, feature extraction as well as different parameters such as the percentage of eyelid closure (PERCLOS), the mouth aspect ratio (MAR) and the eye aspect ratio (EAR).