Research on Small Target Detection Method for Industrial Safety Helmets Based on Improved YOLOv8
Journal of Computing and Information Technology · 27.02.2024 · 31 (2) · 123-136
DOI: 10.20532/cit.2023.1005754 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Computing and Information Technology, online erschienen am 27.02.2024. Verfasst von Lan Yuanshuai und 4 weiteren Autor:innen. Die Arbeit wurde bisher 3-mal zitiert (Crossref).
Abstract (Originalsprache)
Industrial safety helmets are crucial personal protective gear but detecting them as small targets in complex environments is challenging. This work proposes enhancements to the YOLOv8 object detection framework, specifically incorporating a spatial-to-depth (SPD) convolution module and a large selective kernel network (LSKNet). SPD-Conv combines spatial-to- depth layers and non-strided convolutions to retain fine-grained information when downscaling feature maps, while LSKNet introduces dynamic spatial selection and attention for refined context modeling. Our customized model is trained on a dataset of construction hardhat images captured via drones. Quantitative results showcase higher precision and recall over baseline YOLOv8, surpassing competing YOLOv5 versions. An optimized final model outcomes demonstrate accuracy exceeding 90% validation in mAP metric after 200 training rounds. By tackling limitations posed by small, obscured industrial safety gears, this enhanced real-time detection approach provides indispensable technological support for bolstering workplace hazard identification and prevention.
Zitieren
APA
Yuanshuai, L., Mo, C., Chuan, L., Qian, W., & Min, L. (2024). Research on Small Target Detection Method for Industrial Safety Helmets Based on Improved YOLOv8. Journal of Computing and Information Technology, 31(2), 123-136. https://doi.org/10.20532/cit.2023.1005754
Vancouver
Yuanshuai L, Mo C, Chuan L, Qian W, Min L. Research on Small Target Detection Method for Industrial Safety Helmets Based on Improved YOLOv8. Journal of Computing and Information Technology. 2024;31(2):123-136. doi:10.20532/cit.2023.1005754
BibTeX
@article{Yuanshuai2024,
author = {Yuanshuai, Lan and Mo, Chen and Chuan, Li and Qian, Wang and Min, Liao},
title = {Research on Small Target Detection Method for Industrial Safety Helmets Based on Improved YOLOv8},
journal = {Journal of Computing and Information Technology},
year = {2024},
volume = {31},
number = {2},
pages = {123--136},
doi = {10.20532/cit.2023.1005754}
}
EUVIMED