Citation

Khoo, Estelle Hui Jiuan and Ong, Lee Yeng and Leow, Meng Chew (2026) Motorcyclist Helmet Detection with YOLOv8 and CNN. JOIV : International Journal on Informatics Visualization, 10 (4). p. 1886. ISSN 2549-9610

Text Microsoft Word - Estelle 4186-AAP.pdf - Published Version Restricted to Repository staff only Download (4MB)

Abstract

Helmets are crucial for safety, but many motorcyclists still choose not to wear helmets. Hence, enforcing helmet regulations is important, but a huge workforce might be required. With the helmet detection system, enforcing of helmet regulations could be easier. The helmet detection system can identify motorcyclists not wearing helmets, with the captured images serving as evidence for law enforcement. A time-efficient motorcyclist helmet detection could help to give a more real time result, so that the law enforcement could be done with a minimum delay. A two-stage approach is commonly used in motorcyclist helmet detection. The first stage is to extract the ROI (Region of Interest), and the second stage is to detect or classify the extracted ROI. In this study, the pre-trained model, YOLOv8 (You Only Look Once version 8), which is known for its fast and accurate performance, is used for detection task in the first stage. Subsequently, CNN (Convolutional Neural Network) is used for classification task in the second stage. However, most of the existing work only focus on the time efficiency of the detection stage but overlooks the importance of classification time in the second stage. When real-time results are required, classification time significantly affects the overall processing time, especially when multiple ROIs are detected. Therefore, the prediction time for the classification stage is observed, and CNN models with varying numbers of convolution layers and pooling layers are investigated to achieve a shorter classification time while maintaining their accuracy.

Item Type:	Article
Uncontrolled Keywords:	Motorcyclist, helmet, CNN, YOLO
Subjects:	Q Science > QA Mathematics > QA71-90 Instruments and machines
Divisions:	Faculty of Information Science and Technology (FIST)
Depositing User:	Ms Rosnani Abd Wahab
Date Deposited:	05 Oct 2026 02:13
Last Modified:	05 Oct 2026 02:13
URII:	http://shdl.mmu.edu.my/id/eprint/16858

Downloads

Edit (login required)