INTERNATIONAL JOURNAL OF ADVANCE RESEARCH
IN MULTIDISCIPLINARY

Vigilance monitoring plays a crucial role in ensuring safety and efficiency across various industries, including transportation & automotive, security & surveillance, and healthcare & patient monitoring. This project leverages deep learning techniques and computer vision to develop an intelligent drowsiness monitoring system that continuously monitors alertness levels using facial landmarks and eye movement analysis.

The system utilizes the Dlib library for real-time face detection and facial landmark extraction, identifying 68 key facial points. To measure eye openness, the Eye Aspect Ratio (EAR) is computed based on the vertical and horizontal eye distances, helping determine signs of fatigue. The framework integrates OpenCV (cv2) for image processing, imutils for facial alignment, and Scipy for numerical computations. A frame-based counting mechanism tracks continuous eye closure, and when a predefined threshold is exceeded, the system triggers an audible alert using Winsound to prevent potential hazards.

This vigilance monitoring system is designed for real-world applications, ensuring improved safety in driver monitoring systems, security personnel tracking, and healthcare environments where fatigue-related incidents can be life-threatening. By integrating deep learning with facial analysis techniques, the proposed system enhances proactive fatigue detection, offering a low-cost, efficient, and scalable solution for real-time alertness assessment.