Deteksi Helm dan Plat Sepeda Motor Menggunakan YOLOv8 untuk Pencatatan Pelanggar Lalu Lintas

Abstract

The increasing of mobility of the society demanding an effective approach to improve the road safety, mainly for motorcyclists. The high rate of traffic violations and frequent fatal accident have become a serious concern. In this case, computer vision technology, specifically object detection, offers a significant potential to address these safety challenges. This study aims to develop a helmet and motorcycle license plate detection system using YOLOv8 method, it is known for it’s good speed and accuracy. The system was implemented using Python programming language and Flask framework. The test results shows that the system can detect helmets with an accuracy of 80,28%, motorcyclists with an accuracy of 90,26%, and motorcycle license plates with an accuracy of 76,8%. However, the accuracy of license plate extraction is still very low, at 27,5%. Despite this, the system has performed well in detecting helmets, motorcyclists, and motorcycle license plates, but there is still a way to improve the system especially for the license plate extraction. This research is expected to contribute to creating a safer and more organized traffic environment.