EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

VGDS-PointNet++ for organ segmentation and phenotypic trait estimation in greenhouse tomato seedlings

Feiyi Wang, Tianyu Zhu, Lyuwen Huang, Ce Feng, Xi Lu, Wenchong Min, Zheng Wang, Xiaohui Hu, Yanming Nie

Frontiers in Plant Science · 13.04.2026 · 17

DOI: 10.3389/fpls.2026.1753706 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Frontiers in Plant Science, online erschienen am 13.04.2026. Verfasst von Feiyi Wang und 8 weiteren Autor:innen. Beteiligte Einrichtung: College of Information Engineering, Northwest A&F University, Xianyang, Shaanxi, China. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 40 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Introduction

To accurately segment point clouds and quickly calculate leaf length and stem diameter, thereby enabling phenotypic analysis and variety selection of greenhouse tomato plants, this paper proposes a voxel grid downsampling (VGDS)–PointNet++-based model for point cloud segmentation and trait calculation.

Methods

The point clouds of the tomato canopy were acquired using a depth camera. After labeling, point cloud augmentation was performed, and the tomato point cloud dataset (TPCD) containing 1,552 sets of data was rebuilt. Voxel grid downsampling was applied to replace the original sampling strategy of PointNet++. Models of PointNet, PointNet++, VGDS-PointNet++, and Point Transformer were trained with the TPCD and compared on segmentation quality with accuracy and mean Intersection over Union (mIoU). After segmentation, skeletal morphology was fitted for non-occluded leaves by applying a series of surface fitting techniques. The leaf lengths and stem diameters were automatically calculated and compared with the manually measured values.

Results

The validation results showed that the average runtime of voxel grid downsampling was 0.132 s, which was lower than under the same number of sampled points. Compared to the other three models, the proposed model had higher accuracy and mIoU, reaching up to 96.80% and 88.95%, respectively. The proposal’s accuracy and mIoU increased by 3.9% and 4.45% over PointNet++, respectively. The determination coefficient R 2 between the automatic calculation and manual measurement values of leaf length and stem diameter was 0.93 and 0.87, respectively.

Discussion

This can help extract phenotypic traits of tomatoes using depth cameras.

Schlagwörter

Stichwörter der Autor:innen

VGDS-PointNet++leaf length calculationorgan segmentationspace curve fittingtomato canopy

Autor:innen und Institutionen

  • College of Information Engineering, Northwest A&F University, Xianyang, Shaanxi, China.
  • Key Laboratory of Agricultural Internet of Things, Ministry of Agriculture and Rural Affairs, Xianyang, Shaanxi, China.
  • Shaanxi Engineering Research Center of Agricultural Information Intelligent Perception and Analysis, Northwest A&F University, Yangling, China.
  • College of Horticulture, Northwest A&F University, Xianyang, Shaanxi, China.

Zitieren

APA

Wang, F., Zhu, T., Huang, L., Feng, C., Lu, X., Min, W., Wang, Z., Hu, X., & Nie, Y. (2026). VGDS-PointNet++ for organ segmentation and phenotypic trait estimation in greenhouse tomato seedlings. Frontiers in Plant Science, 17. https://doi.org/10.3389/fpls.2026.1753706

Vancouver

Wang F, Zhu T, Huang L, Feng C, Lu X, Min W, et al. VGDS-PointNet++ for organ segmentation and phenotypic trait estimation in greenhouse tomato seedlings. Frontiers in Plant Science. 2026;17. doi:10.3389/fpls.2026.1753706

BibTeX
@article{Wang2026,
  author = {Wang, Feiyi and Zhu, Tianyu and Huang, Lyuwen and Feng, Ce and Lu, Xi and Min, Wenchong and Wang, Zheng and Hu, Xiaohui and Nie, Yanming},
  title = {VGDS-PointNet++ for organ segmentation and phenotypic trait estimation in greenhouse tomato seedlings},
  journal = {Frontiers in Plant Science},
  year = {2026},
  volume = {17},
  doi = {10.3389/fpls.2026.1753706}
}
Kostenlos starten →