Automatic Driver Alertness Detection using Machine Learning

Nowadays driver drowsiness is a major cause of road accidents. Drowsiness is the condition in which the driver is in between the sleepy and wake-up state. Besides all these causes of the accident, this research aims to give the solution to this problem. So, to avoid road accidents, it is imperative to develop a system for detecting the symptoms of drowsiness well on time. The detected symptoms could be physical or psychological signs. The primary goal of our research is to figure out how to prepare our dlib shape indicator and how to utilize it on facial pictures and real-time video. The focused area in this research is the eye region the system detects the eye portion and subsequently calculates the eye aspect ratio. After that, the system makes an alert if the eyes could not find an active state within 36 consecutive frames. As soon as, the eyes are found in an inactive state, the alarm is provoked. This alert makes the driver attentive again. To accomplish the objective, we develop the application in open CV using python language, the result shows high detection of driver drowsiness with 92.57% accuracy.