ASAP: anti-sleep alarm and prompter system using images processing for drowsy drivers

Abstract

As the number of accidents due to driver’s drowsiness grows significantly, demand for drowsiness detection and anti-collision with automatic braking mechanism arises. In this research, the proponents developed an anti-sleep alarm and prompter system for drowsy drivers. The system comes with an automatic braking mechanism that pulls the physical brake pedal of the car when drowsiness or an obstacle is detected. Facial recognition, object detection, and the automatic braking mechanism of the prototype are performed and controlled by using a Raspberry Pi microprocessor and an Arduino Uno microcontroller. In drowsiness detection test drives, the automated braking system activated when the system detected that the driver’s eyes were closed for more than 3 seconds above the EAR trigger value of 0.30. In anti-collision test drives, the automatic braking system activated between 10 to 5 meters from the obstacle.