IOT Based Real-Time Drowsy Driving Detection System for the Prevention of Road Accidents

• Main Authors: Hossain, Md. Yousuf, George, Fabian Parsia
• Format: Conference Proceeding
• Language: English
• Subjects: Cameras; Computer science; Computer Vision; Drowsy Driving; Ear; Eye Aspect Ratio; Face; Facial Landmark; Feature extraction; Pi Camera Module; Raspberry Pi; Training; Vehicles
• ISSN: 2189-8723
• DOI: 10.1109/ICIIBMS.2018.8550026

At present time, drowsy driving has become one of the major issues of the traffic collision. According to statistics, a large number of road accidents occur due to drowsy driving which results in severe injuries and deaths. For this reason, various studies were done in designing systems that can examine the driver fatigue and alert him beforehand, thus preventing him to fall asleep behind the wheel and cause an accident. Some traditional approaches used vehicle-based measures to design their system, however, such measurements are highly influenced by the structure of the road, type of vehicle and the driving skill. Other approaches used psychological measures for their system that tend to provide better accuracy in monitoring the drowsiness of the driver. However, such techniques are usually intrusive as electrodes are required to be placed on the head and body. Furthermore, there are few existing researches in which subjective measurements are used as the input for the system, but, such methods can distract the driver and lead to an ambiguous result. In this paper, we proposed a system that is absolutely nonintrusive and real-time. Our proposed system used the eye closure ratio as input parameter to detect the drowsiness of the driver. If the eye closure ratio deteriorates from the standard ratio, the driver is alerted with the help of a buzzer. For our system, a Pi camera is used to capture the images of the driver's eye and the entire system is incorporated using Raspberry-Pi.