Real-Time Motorbike Helmet Detection System Using Hybrid Deep Learning Model

DOI

10.2991/978-94-6239-664-7_40How to use a DOI?

Keywords

Motorcycle safety; Helmet detection; Deep learning; YOLOv8; ResNet50

Abstract

In Bangladesh, motorcycle accidents are a major cause of mortality; most deaths are ascribed to riders not wearing helmets. This study intends to use a hybrid deep learning model to construct an effective, real-time helmet detection system in response to this rising issue. The system is intended to recognize helmets, riders, and license plates in intricate road scenarios by integrating YOLOv8 for object identification, ResNet50 for rider classification, and Paddle OCR for text recognition. The study used a large dataset from the Kaggle Helmet Detection Dataset and Bangladeshi real-life video footage, preprocessed using Roboflow for cleaning, annotation, and data augmentation. The model was trained using multiple checkpoints and selected best.pt due to superior performance. Data exploration and visualization techniques were used to ensure the dataset’s quality and balance. Feature extraction techniques like edge detection and color histograms improved the model’s accuracy in distinguishing helmets and riders. With a precision of 0.929, recall of 0.932, and mean average accuracy (mAP) of 0.949 at 50 percent IoU, the model’s preliminary evaluations yield encouraging findings. These outcomes show how well the technology detects helmets in real-time, even in intricate traffic situations. In addition to providing insights that may be used to guide policy choices targeted at lowering motorcycle deaths, the project’s results highlight the significance of incorporating deep learning technology for traffic safety.

Copyright

© 2026 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Ahnaf Tahmid Jamee
AU  - Md Abdullah
AU  - A. K. M. Sakibul Alam Adib
AU  - Mohammad Abdullah
AU  - Nowshad Ahamed
AU  - Tahsan Mahmood Tariq
AU  - MD. Tahmeed Kowsher Hameem
AU  - Tariqul Islam
PY  - 2026
DA  - 2026/06/08
TI  - Real-Time Motorbike Helmet Detection System Using Hybrid Deep Learning Model
BT  - Proceedings of the International Conference on Intelligent Data Analysis and Applications (IDAA 2025)
PB  - Atlantis Press
SP  - 581
EP  - 592
SN  - 1951-6851
UR  - https://doi.org/10.2991/978-94-6239-664-7_40
DO  - 10.2991/978-94-6239-664-7_40
ID  - Jamee2026
ER  -