AI-Based System for Detecting Driver Drowsiness

Abstract

This project aims to develop an advanced Driver Drowsiness Detection system that utilizes real-time monitoring of a driver's eye movements to identify signs of fatigue. The system employs a camera mounted within the vehicle to track the driver's face and focus on detecting prolonged eye closures. Leveraging deep learning techniques with a Convolutional Neural Network (CNN) built-in Keras, the system classifies eye states (open or closed) in real-time. Facial landmark identification is achieved through OpenCV, while adaptive lighting compensation ensures consistent performance under varying light conditions. The system incorporates an advanced alert mechanism, combining audible alarms with vehicle-integrated haptic feedback to prompt immediate driver response.

Additionally, it features real-time calibration capabilities to adapt to individual driver traits, enhancing detection accuracy. The non-invasive nature of this technology eliminates the need for wearable devices, relying solely on the camera and software for operation. This innovative solution significantly improves road safety by reducing accidents caused by fatigue and distraction. Its versatility allows for deployment in various vehicles, catering to both personal and commercial applications. By issuing early warnings, the system keeps drivers vigilant, contributes to smarter vehicular systems, and paves the way for safer roads, ultimately saving lives.