Deep Learning Based Driver Drowsiness Monitoring and Alert System

Road traffic accidents destroyed the lives of over 1.35 million people every year worldwide. When compared to an attentive driver, a distracted driver has significantly longer reaction times. This fact contributes to the risk of such drivers becoming involved in an accident. Driver monitoring system is found to lower the risk of accidents caused by tired, sleepy or distracted drivers. To notify drivers and reduce the large incidence of collisions, numerous detecting systems have been created. But most of them depend on specialised hardware that is integrated inside the car. Accordingly, this research aims at designing of a novel effective Human Machine Interface (HMI) for driver monitoring system using deep learning. This technology works by monitoring the driver's eye movement and notifying the driver by triggering the buzzer when he/she is fatigued. In addition, the system also has an option for reducing the gear motor speed automatically when drowsiness is detected. The deep learning algorithm employed in this work is Convolutional Neural Network (CNN), which is utilized for alert generation, image classification, feature extraction and detection. The system continually detects the driver's eyelid movements and if drowsiness is identified, the gadget informs him with a buzzer sound alarm. The suggested method is found to be stable and reliable for face and eye tracking.