Improvement of the keypoint detection network for tobacco leaves based on the YOLOv8n-pose framework
IET Conference Proceedings · 01.08.2026 · 2025 (62) · 309-315
DOI: 10.1049/icp.2026.2653 ↗Kurz gefasst
Fachartikel in der Zeitschrift IET Conference Proceedings, erschienen am 01.08.2026. Verfasst von Liming Sui und 3 weiteren Autor:innen.
Abstract (Originalsprache)
In response to the demands of agricultural modernization and the tobacco market, with the aim of enhancing the quality and efficiency of tobacco leaf harvesting, this study focuses on the vision system of tobacco leaf picking robots and conducts research on key point detection and localization technologies for tobacco leaves. A dataset of tobacco leaf images was established by collecting image samples and employing data annotation and augmentation techniques. To enhance feature extraction capability and computational efficiency, an improved YOLOv8-Pose network model was developed by optimizing the backbone, neck, and head architectures within the YOLOv8-Pose framework. Ablation, comparative, and visualization experiments validated the effectiveness of the proposed improvement strategy. The experimental results indicate that the enhanced model achieved a 7% increase in precision and a 3.2% improvement in mean average precision compared to the baseline model, while reducing the number of parameters and floating-point operations by 43.2% and 39.8%, respectively.
Zitieren
APA
Sui, L., Gao, C., Zhang, Y., & Tan, D. (2026). Improvement of the keypoint detection network for tobacco leaves based on the YOLOv8n-pose framework. IET Conference Proceedings, 2025(62), 309-315. https://doi.org/10.1049/icp.2026.2653
Vancouver
Sui L, Gao C, Zhang Y, Tan D. Improvement of the keypoint detection network for tobacco leaves based on the YOLOv8n-pose framework. IET Conference Proceedings. 2026;2025(62):309-315. doi:10.1049/icp.2026.2653
BibTeX
@article{Sui2026,
author = {Sui, Liming and Gao, Chenhang and Zhang, Yin and Tan, Dingzhong},
title = {Improvement of the keypoint detection network for tobacco leaves based on the YOLOv8n-pose framework},
journal = {IET Conference Proceedings},
year = {2026},
volume = {2025},
number = {62},
pages = {309--315},
doi = {10.1049/icp.2026.2653}
}
EUVIMED