Detecting Driver Fatigue in Real Time: An OpenCV and Keras Implementation

It is vital to address the major public health concern of sleep-related driving. According to recent study, drunk drivers are responsible for 20% of all automotive accidents. When individuals do not get enough sleep, they are more likely to fall asleep, increasing the risk of a traffic accident, especially for drivers. The development of more advanced driver assistance systems is increasingly focused on improving the reliability of fatigue identification. To prevent road accidents, the current program aims to develop a technology that can identify driver drowsiness. As a starting point, it was considered that photos of the driver with their eyes closed represented sleepiness while images with their eyes open represented attentiveness. After identifying the face using Naïve Bayes Region of Interest, isolate the lips and eyes. The driver's face is visible in a snapshot taken inside a car. And it states that the driver is tired in voice. A camera often shoots photos in RGB format (Red, Green, and Blue). It also gives suggestions for neighbouring hotels.