SISTEM DETEKSI PENGENDARA MOTOR MENGGUNAKAN HELM DAN TIDAK MENGGUNAKAN HELM BERBASIS ALGORITMA YOU ONLY LOOK ONCE (YOLO)

ANDANTITO, MIRFAK (2023) SISTEM DETEKSI PENGENDARA MOTOR MENGGUNAKAN HELM DAN TIDAK MENGGUNAKAN HELM BERBASIS ALGORITMA YOU ONLY LOOK ONCE (YOLO). Tugas Akhir thesis, University of Technology Yogyakarta.

Text 5190411023_mirfak ardiantito ABSTRAK.pdf Download (15kB)

Abstract

ABSTRACT Traffic is one of the most challenging and difficult problems in city management, especially in developing countries, the death rate due to accidents is also getting higher every year, especially for motorcyclists who violate the rules that have been determined. In order to ensure traffic safety, researchers create artificial intelligence such as object detection, which can make it easier for researchers to recognize objects that will become research material. There are still relatively few studies that discuss helmet detection using the YOLO method, mostly this method is used to detect vehicles and vehicle plates. The author tries to use the YOLO method to detect motorcycle helmets in Indonesia, the purpose of the study is to be able to distinguish motorcyclists using helmets and not using helmets using the You Only Look Once (YOLO) algorithm through images, and realtime. This research uses a dataset of 428 images with two classes, namely helmets and heads, training data is done with 21 filters, 64 batches, 16 subdivisions, and 5000 max batches. The test results obtained the highest confidence value in the helmet class of 0.96, the non_helm class of 0.91, and the mAP value obtained was 93.12%. This shows that the You Only Look Once (YOLO) algorithm can recognize helmet and non_helmet objects in video and realtime using pre-trained weights that have been trained by themselves. Keywords: Helmet, Machine Learning, Object Detection, Character Recognition, Deep Learning, You Only Look Once (YOLO).

Item Type:	Thesis (Skripsi, Tugas Akhir or Kerja Praktek) (Tugas Akhir)
Subjects:	T Technology > T Technology (General)
Divisions:	Fakultas Sains Dan Teknologi > S1 Informatika
Depositing User:	Kaprodi S1 Informatika UTY
Date Deposited:	03 Oct 2023 03:57
Last Modified:	03 Oct 2023 03:57
URI:	http://eprints.uty.ac.id/id/eprint/13648

Actions (login required)

View Item