Drowsiness Detection Using Deep Learning and Camera-Based System

Abstract

To address this issue, we propose a drowsiness detection system that leverages deep learning techniques, specifically a Convolutional Neural Network (CNN), to monitor and alert drivers in real-time. Our system utilizes readily available hardware components, including a camera, USB cable, laptop as the processing unit, and a speaker as a warning system. The camera captures live video feed of the driver's face, which is then processed by the CNN model in real-time. The CNN model is trained to analyse facial features and detect signs of drowsiness such as drooping eyelids, yawning, and head nodding. When the CNN model identifies drowsiness indicators in the driver's facial expressions, it triggers an alert system. The alert system, through the speaker, emits warning sounds or voice prompts to grab the driver's attention and prompt them to stay alert. The USB cable connects the camera to the laptop, ensuring a seamless data transfer and processing flow. This drowsiness detection system not only enhances road safety but also offers the advantage of easy integration into existing vehicles, as it relies on commonly available components. Our research demonstrates the feasibility of implementing a cost-effective and efficient solution to mitigate the dangers of drowsy driving, ultimately saving lives on the road