Real Time Exhaustion Sensing for Motor Vehicle Operators

Drowsy driving remains a major global safety concern, contributing to a substantial number of road accidents. To address this critical issue, this paper proposes an innovative Image Recognition-Based Drowsiness Detection System leveraging advanced computer vision and machine learning techniques. The system operates in real-time, continuously monitoring the driver’s facial expressions and, most critically, their eye movements within the vehicle cabin. It employs sophisticated Eye Aspect Ratio (EAR) tracking algorithms based on Dlib facial landmarks to extract crucial features, including the eye closure duration. These cues are processed by a robust rule-based classification model that identifies sustained eye closure indicative of fatigue. Upon detecting signs of drowsiness, the system issues a timely, non- intrusive audible alert, prompting the driver to take corrective action. By providing a reliable and cost-effective driver alert mechanism with a demonstrated 95.2% detection accuracy, this research contributes significantly to mitigating drowsy-driving related accidents and promoting safer roads for all.