Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Artículo Open Access |
| Published: |
Wiley
2025
|
| Subjects: | |
| Online Access: | https://onlinelibrary.wiley.com/doi/10.1002/tee.70100 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Table of Contents:
- A Lightweight Method for Transmission Line Component Detection Based on YOLOv8n Guiwen Lan Zhan Zhong Ruidong Guo Hanqiang Huang Zirui Xu Xinyue Ren IEEJ Transactions on Electrical and Electronic Engineering Many existing methods for transmission line component detection are unfeasible to deploy on resource‐constrained devices. In this paper, an improved YOLOv8n for transmission line component detection is proposed to make a balance between detection efficiency and model size by using many lightweight blocks in the YOLOv8n's neck. To pay more attention to the information of interest so as to reduce computational load, an EMA module and a lightweight RepViT block are embedded in the C2f module of the 12th layer of the neck network. A C2f_SCConv module is designed and used to replace the C2f module in the 18th layer of the neck network in order to improve feature fusion across different layers. The C2f modules in the 15th and 21st layers of the neck network are replaced with a C2f_Faster module respectively, meanwhile maintaining accuracy and reducing the number of convolutional layers to improve detection speed. The Wise‐IoU loss function is chosen to make the improved model more flexible and robust in object detection. Experimental results show that the proposed model obtains a significant increase in detection accuracy with a lightweight model size. © 2025 Institute of Electrical Engineers of Japan and Wiley Periodicals LLC. 10.1002/tee.70100 http://onlinelibrary.wiley.com/termsAndConditions#vor