Canopy and leaf vein segmentation and topology-guided localization of stem points in tobacco plants using YOLO-ODS
4 affiliations
- 1College of Engineering, South China Agricultural University
- 2Key Lab of Key Technology for South Agricultural Machine and Equipment, Ministry of Education, P.R. China (KTSAME)
- 3College of Food Science, South China Agricultural University
- 4School of Information Technology and Engineering, Guangzhou College of Commerce
Abstract
Accurate stem point localization is critical for intelligent tobacco production but remains challenging in low-altitude nadir-view UAV imagery. This is because the visible upper stem after topping is small, weakly textured, and visually similar to surrounding canopy tissues. This study formulates stem point localization as a topology-guided geometric inference problem based on canopy and leaf vein structures. A lightweight instance segmentation model, YOLO-ODS, was developed on YOLOv11n-Seg to jointly segment tobacco canopies and leaf veins from UAV imagery. The model integrates an Orientation-Aware Convolution Module (OACM), a Dual Interactive Attention Module (DIAM), and Slide Loss to enhance directional feature representation and improve discrimination of fine-grained structures. Based on segmentation results, leaf vein skeletons are extracted and local trend lines are fitted to estimate stem points via geometric convergence under canopy constraints. Experimental results showed that YOLO-ODS achieved a Recall of 86.79%, a mask [email protected] of 90.19%, and a mask [email protected]:0.95 of 67.20%, outperforming the YOLOv11n-Seg baseline by 2.20, 1.14, and 2.10 percentage points, respectively. For the leaf vein category, the mask [email protected]:0.95 reached 51.49%. With 2.86 M parameters and 10.19 GFLOPs, YOLO-ODS remained lightweight for field deployment. The proposed localization method achieved an MAE of 12.05 px and an RMSE of 12.99 px, reducing MAE and RMSE by 51.2% and 52.6%, respectively, compared with YOLOv11n-pose. These results suggest that integrating lightweight instance segmentation with leaf vein topology may provide a promising approach for improving stem point localization under complex field conditions.
Related
- Revisiting genome-guided RNA localization in bacteria
- Agentic AI enables cell segmentation-free cell-type inference and cell-type-guided cell segmentation in spatial transcriptomics
- A public-data reproducibility benchmark of a 2.5-D U-Net with a slice-level topology term for type-B aortic dissection segmentation
- MCseg: AI agent-guided workflow search for no-code cell segmentation and transcript attribution in spatial transcriptomics
- TAL-CGBR: A class-guided boundary regression approach for Temporal Action Localization
BibTeX
@article{qing2026canopy,
title = {{Canopy and leaf vein segmentation and topology-guided localization of stem points in tobacco plants using YOLO-ODS}},
author = {Jiaxing Qing and Donghao Li and Xiwen Luo and Zhihao Zeng and Jinhong Chen and Yawen Liu and Jingqi Xu and Shan Zeng and Yukun Yang and Yu Wang},
journal = {Frontiers in Plant Science},
year = {2026},
month = oct,
doi = {10.3389/fpls.2026.1930225},
url = {https://www.frontiersin.org/articles/10.3389/fpls.2026.1930225/full},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Canopy and leaf vein segmentation and topology-guided localization of stem points in tobacco plants using YOLO-ODS
AU - Jiaxing Qing
AU - Donghao Li
AU - Xiwen Luo
AU - Zhihao Zeng
AU - Jinhong Chen
AU - Yawen Liu
AU - Jingqi Xu
AU - Shan Zeng
AU - Yukun Yang
AU - Yu Wang
PY - 2026
DA - 2026/10/07
JO - Frontiers in Plant Science
DO - 10.3389/fpls.2026.1930225
UR - https://www.frontiersin.org/articles/10.3389/fpls.2026.1930225/full
AB - Accurate stem point localization is critical for intelligent tobacco production but remains challenging in low-altitude nadir-view UAV imagery. This is because the visible upper stem after topping is small, weakly textured, and visually similar to surrounding canopy tissues. This study formulates stem point localization as a topology-guided geometric inference problem based on canopy and leaf vein structures. A lightweight instance segmentation model, YOLO-ODS, was developed on YOLOv11n-Seg to jointly segment tobacco canopies and leaf veins from UAV imagery. The model integrates an Orientation-Aware Convolution Module (OACM), a Dual Interactive Attention Module (DIAM), and Slide Loss to enhance directional feature representation and improve discrimination of fine-grained structures. Based on segmentation results, leaf vein skeletons are extracted and local trend lines are fitted to estimate stem points via geometric convergence under canopy constraints. Experimental results showed that YOLO-ODS achieved a Recall of 86.79%, a mask [email protected] of 90.19%, and a mask [email protected]:0.95 of 67.20%, outperforming the YOLOv11n-Seg baseline by 2.20, 1.14, and 2.10 percentage points, respectively. For the leaf vein category, the mask [email protected]:0.95 reached 51.49%. With 2.86 M parameters and 10.19 GFLOPs, YOLO-ODS remained lightweight for field deployment. The proposed localization method achieved an MAE of 12.05 px and an RMSE of 12.99 px, reducing MAE and RMSE by 51.2% and 52.6%, respectively, compared with YOLOv11n-pose. These results suggest that integrating lightweight instance segmentation with leaf vein topology may provide a promising approach for improving stem point localization under complex field conditions.
SN - 1664-462X
ER -