Recent progress in computer vision has made it possible to support drivers through technologies such as automatic and self-driving cars. However, driver fatigue and drowsiness still account for nearly 20% of road accidents, making it a serious safety concern. Several solutions have been proposed to address this issue, but many of them are not well suited for real-time applications. One of the main challenges faced by these systems is maintaining reliability under variations in human facial features and changing lighting conditions. To address these challenges, this work proposes an intelligent monitoring system aimed at significantly reducing road accidents caused by driver drowsiness. The system analyzes key facial characteristics of the driver, including eye closure percentage, eye-mouth aspect ratio, blink rate, yawning patterns, and head movements. A webcam continuously monitors the driver, and Haar cascade classifiers are used to detect the driver’s face and eyes. The detected eye images are then processed using a custom-designed Convolutional Neural Network (CNN) to determine whether the left and right eyes are closed. Based on this classification, an eye closure score is calculated. If the system detects signs of drowsiness, an alarm is immediately triggered to alert the driver.
B.CHARAN et al. (Sun,) studied this question.