A Machine Learning Approach to Driver Fatigue Detection via Facial Behavior and SVM

Abstract

This project focuses on detecting driver drowsiness by analyzing the driver's visual behavior using a webcam and a machine learning model based on the Support Vector Machine (SVM) algorithm. The system utilizes the built-in webcam to continuously capture images of the driver, and with the help of OpenCV, it extracts key facial features such as eye and mouth positions.A pre-trained SVM model is employed to classify signs of drowsiness. The system monitors eye closure over a sequence of frames—specifically, detecting if the driver's eyes remain closed for 20 consecutive frames—or signs of yawning based on mouth movements. To achieve this, the application calculates the Euclidean distance between facial landmarks related to the eyes and mouth. If these distances align with drowsiness indicators, the system triggers an alert message to warn the driver.This real-time monitoring approach is designed to reduce the risk of accidents caused by driver fatigue, offering a practical and efficient solution for enhancing road safety.