Real Time Alert System for Drivers

The significant causes of road accidents are fatigue and drowsiness of drivers. The usage of a driver drowsiness system will help in reducing the accidents which occur by driver’s tiredness and resulting in more enhanced vehicular safety. This paper helps in presenting a system which can detect the drowsiness using facial movements such as eye blinks and yawns and alerting the driver about possible accidents using Raspberry Pi. The system will monitor the driver’s frame continuously with the help of a Raspberry Pi camera. Then the captured frame is processed first using Viola Jones Algorithm to detect the face, eyes and mouth and eventually drowsiness and yawning. In the system, Eye Aspect Ratio (EAR) is calculated with the help of facial coordinates to detect the drowsiness. And yawn value (distance between upper lip and lower lip) is calculated to detect the yawning. And if drowsiness or yawning is discovered then an audio alarm using the “simple-audio” library in the python will be produced. System is tested with more than 25 videos and generating audio alert as expected.