Real-Time Driver Drowsiness Detection Using Eye Aspect Ratio and Facial Landmark Analysis

Abstract

Driver drowsiness detection is crucial to preventing road accidents caused by fatigue. This paper proposes a non-intrusive, real-time system based on facial landmark detection and the Eye Aspect Ratio (EAR) using a standard webcam. The system continuously monitors eye activity and triggers an alert when signs of drowsiness are detected, measured by sustained low EAR values. The approach integrates a pre-trained face detector, facial landmark predictor, and an EAR-based thresholding mechanism to determine eye closure. High accuracy is demonstrated by the experimental results in detecting drowsiness, making the system suitable for embedded or mobile deployment in automotive applications.