RANCANG BANGUN ALAT PENDETEKSI KANTUK PADA KENDARAAN RODA EMPAT DENGAN METODE HAAR CASCADE

Abstract

Based on the Surabaya Police Traffic Unit, the number of accidents due to traffic violations in Surabaya from January to August 2019 reached 882. One of the biggest causes of accidents is caused by human error. One of the human error factors is fatigue or sleepiness. Sleepiness is a very common thing that happens to everyone. This can be caused by various factors, namely fatigue, lack of sleep, overeating. Drowsiness can be defined as a process produced by circadian rhythms and the need for sleep. In a drowsy state, a person can increase the blink of an eye as much as 20% of the frequency of blinks per minute. In addition, a person experiences microsleep with a duration of eye closure of 0.5 seconds or more. The eye recognition process carried out by computer vision is not as easy as what is done by humans directly. While humans themselves are very easy to recognize someone very quickly without having to think long. While computer vision is very slow in recognition. So in this research, a sleep detection device with facial recognition will be made which detects sleepiness in the eyes for safety when driving on the highway so that it can reduce the risk of accidents on the highway, using a raspberry pi microcontroller using the Haar Cascade method which requires very fast eye recognition. It is hoped that from my research this tool can reduce the number of accidents in Indonesia, especially in East Java, the city of Surabaya.