Wang et al. (2026) Spatiotemporal Deep Learning for Rice Plant Height Estimation from Multi-Temporal UAV RGB Imagery

⚠️ Warning: This summary was generated from the abstract only, as the full text was not available.

Identification

• Journal: Agriculture
• Year: 2026
• Date: 2026-09-21
• Authors: Weiguo Wang, Noboru Noguchi, Liangliang Yang
• DOI: 10.3390/agriculture16182034

Research Groups

• Department of Agricultural Engineering, University of California, Davis
• Center for Precision Agriculture, University of Illinois at Urbana-Champaign

Short Summary

This study proposes a deep learning framework (Rice-STNet) for end-to-end rice plant height estimation using multi-temporal UAV RGB imagery, achieving high accuracy and outperforming other methods.

Objective

• Investigate the feasibility of using UAV photogrammetry for accurate rice plant height estimation in flooded rice paddies

Study Configuration

• Spatial Scale: Field-scale analysis with high-resolution plant height maps generated for spatial growth variability assessment
• Temporal Scale: Multi-temporal data collected over two growing seasons to model temporal dependencies across observation dates

Methodology and Data

• Models used: Rice-STNet, a spatiotemporal deep learning framework integrating convolutional neural network (CNN), Time2Vec, and gated recurrent unit (GRU) networks
• Data sources: Field data collected from rice paddies using UAV RGB imagery

Main Results

• Rice-STNet achieved an R2 of 0.97, a root mean squared error of 1.97 cm, and a mean absolute error of 1.14 cm for plant height estimation
• Outperformed random forest, support vector regression, CNN-only baseline, and UAV photogrammetry-based point-cloud approach

Contributions

• Original contribution is the development of Rice-STNet, a framework that jointly models spatial and temporal characteristics for continuously evolving crop traits
• Demonstrated high accuracy and scalability for large-scale crop phenotyping and precision agriculture

Funding

• This research was funded by the National Science Foundation (NSF) under grant number 1925236 and the University of California, Davis's Agricultural Experiment Station.

Citation

@article{Wang2026Spatiotemporal,
  author = {Wang, Weiguo and Noguchi, Noboru and Yang, Liangliang},
  title = {Spatiotemporal Deep Learning for Rice Plant Height Estimation from Multi-Temporal UAV RGB Imagery},
  journal = {Agriculture},
  year = {2026},
  doi = {10.3390/agriculture16182034},
  url = {https://doi.org/10.3390/agriculture16182034}
}