Automated Driver Drowsiness Detection System Using Artificial Intelligence

Abstract

This paper presents an AI-based driver drowsiness detection system that monitors driver fatigue in real time using a low-cost webcam. The system detects facial features with OpenCV and classifies eye state using a Convolutional Neural Network (CNN). It also analyses yawning and ocular redness to estimate a weighted drowsiness score. An adaptive calibration process personalises detection for each driver, while graded audible alerts provide timely warnings. To improve energy efficiency, the system supports Manual, Continuous, and Auto operating modes. The proposed approach is low-cost, non-intrusive, and suitable for various vehicles, helping reduce fatigue-related accidents and improve road safety.