Use of Facial Information to Detect Driver's Sleepiness and Disorientation

Road accidents in Iran account for a large volume of injuries and mortality, which have a lot of financial losses for society. Therefore, recognizing driver's sleepiness can have a great effect in reducing these injuries. One of the ways to detect fatigue and distraction is the use of surveillance systems from the driver's face. In this paper, a driver's face monitoring system is designed to estimate the driver's consciousness by extracting signs of tiredness and distraction from the eye area. These characteristics are then processed by the KNN algorithm and the RBF network in order to estimate the amount of distraction of the driver's senses. The results of the experiments on the images in the actual environment show that the proposed method has a very good accuracy. In terms of implementing the algorithm, the accuracy of the proposed system is 95.7%.