HomeAni: Letran Calamba Research Reportvol. 19 no. 1 (2023)

Real-Time Driver Drowsiness Detection through Facial Detection using Viola-Jones Algorithm

Karl Mario Tan | John Matthew Ibalio | Francis Dale Escobin

 

Abstract:

Driver fatigue, also known as drowsy driving, is one of the prime causes of road accidents all over the world. With the constant development of technology today, the researchers thought of using a device such as a mobile phone, that could help the drivers be notified that they are undergoing a state of drowsiness. The system aimed to detect the driver’s face using the applied algorithm called Viola-Jones, which is one of the first object detection frameworks to provide real-time competitive object detection rates. After detecting the driver’s face, the system would get the eyes’ feature to determine if the driver performs eye blink. Using this method, the system would be able to monitor the indications of drowsiness using behavioral basis, specifically, frequent blinking and microsleep. Once the indications exceed the set threshold, the driver would be notified by triggering the alarm, loud music and short messaging system (SMS) that would be sent to the number that the driver inputted and directly store the date and time of drowsy driving. The researchers conducted a survey based on the test with the system which indicated that the system’s functionality, efficiency, portability, reliability and usability had a high overall mean which showed that the system had met the respondent’s expectation. Based on the data gathered from software testing, the system could correctly identify drowsiness using blink rate with an accuracy of ninety-five percent (95%), while microsleep has an accuracy of ninety percent (90%).