Driver Drowsiness Detection System Based on Yawn Detection

Accidents caused by driver fatigue are increasing day by day and are becoming a major problem in recent times. Therefore, designing and developing a system to detect a drowsy driver and give an instant warning when the driver is distracted is a necessary step to prevent accident. Driving is a part of daily life and often leads to negative emotions such as anger or stress that can affect people’s safety and health for a long time. In recent years, the availability of new technology to understand human emotions has increased the interest in incorporating emotion for cars. Real-time emotional intelligence has been an active research area for the past few years. Methods based on yawn detection, drowsiness detection, and driving pattern detection are used to successfully detect and track faces in a video stream by using a mobile application. Yawn detection can be done by using LBP and thresholding. Using this application, driver can be continuously monitored, and if sleep is detected an alarm will get triggered. An accuracy of 97% is obtained through this proposed method by using a dataset of 100 different people.