Drowsiness Detection and Accident Prevention System Using OpenCV and Raspberry Pi

This paper presents a robust solution for driver drowsiness detection and accident prevention using TensorFlow and OpenCV-based computer vision techniques integrated with Raspberry Pi and a suite of peripheral devices. The system utilizes a webcam to continuously monitor the driver's facial expressions and eye movements, employing deep learning algorithms to detect signs of drowsiness such as prolonged eye closure or head nodding. Upon detecting potential drowsiness, the system triggers a multi-layered alert mechanism: a loud buzzer and a flashing red LED inside the vehicle alert the driver immediately. Simultaneously, the system sends an SMS alert via the GSM module to predefined emergency contacts, providing real-time coordinates (latitude and longitude) of the vehicle's location using the integrated GPS module. This comprehensive approach not only aims to mitigate driver fatigue-related accidents but also ensures prompt response and intervention by alerting both the driver and emergency responders effectively. By integrating real-time monitoring, alerting, and location-based communication, this system enhances road safety by proactively addressing hazardous driving conditions caused by driver drowsiness. The effectiveness and reliability of the proposed system are demonstrated through experimental validation, highlighting its potential for widespread adoption in vehicles to prevent accidents and save lives.