Driver Drowsiness Detection Using Python

Abstract

Drowsiness has emerged as a pervasive global concern, proving to be a major factor in catastrophic accidents that result in fatalities and severe injuries. The topic proposes novel experimental model designed to detect driver drowsiness, aiming to mitigate the occurrence of accidents and enhance overall transport safety. The approach integrates two distinct methods for effective drowsiness detection. Firstly, facial recognition techniques are employed to capture the driver's face and perform eye retina detection. Facial features are extracted, and blinking values are calculated. Threshold values for blinking are then established to gauge the driver's level of drowsiness. Secondly, an Arduino modules equipped with force sensors, is integrated into the system. The module continuously monitors the real-time pressure exerted by the driver's hands on the steering wheel. Threshold values for hand pressure are set to determine the driver's engagement level. The decision-making process involves synthesizing the results from both methods to make a comprehensive assessment of the driver's alertness. If either the facial recognition system or the force sensors indicate drowsiness beyond the set thresholds, an alert is triggered. The implementation includes an alert system that provides visual, auditory, or haptic cues to prompt the driver to take corrective action upon detecting drowsiness. This dual-method approach aims to create a robust system for detecting and addressing driver fatigue, ultimately contributing to the reduction of accidents caused by drowsy driving and promoting overall road safety