Vision-Based Drowziness Detection for In-Vehicle Driver Monitoring

Driver drowsiness is a major contributor to road accidents worldwide, posing a serious threat to traffic safety.This paper presents a real-time drowsiness detection system using Python and OpenCV to monitor driver fatigue based on facial landmarks.The proposed method utilizes a webcam to capture live video frames and employs Haar Cascade classifiers and Dlib facial landmark detection to analyze key facial features such as eye aspect ratio (EAR) and mouth movement.A threshold-based approach is used to identify prolonged eye closure and yawning, which are strong indicators of drowsiness.When such symptoms are detected, the system triggers an audible alert to warn the driver.The model is lightweight, cost-effective, and can be integrated into in-vehicle systems for enhanced road safety.Experimental results demonstrate the effectiveness and real-time performance of the system under varied lighting and head movement conditions.