Rice multi-variety yield prediction using multi-temporal fusion of thermal infrared and multispectral UAV data based on random Forest feature selection
- 1College of Agriculture, Rice Industry Technology Research Institute, Guizhou University, Guiyang 550025, PR China
- 2College of Electrical and Information, Key Laboratory of Northeast Smart Agricultural Technology, Ministry of Agriculture and Rural Affairs, Northeast Agricultural University, Harbin 150030, PR China
- 3Institute of Rural Revitalization, Guizhou University, Guiyang 550025, PR China
Abstract
Rice is a major staple crop that feeds nearly half of the global population; thus, its accurate yield estimation is critical for food security and agricultural management. To improve prediction accuracy and applicability, in this study, we propose a multi-temporal fusion framework that integrates unmanned aerial vehicle (UAV)-based multispectral and thermal infrared data with feature selection and deep learning models. UAV data were collected from 57 rice varieties at four key growth stages (tillering, heading, grain filling, and maturity). Vegetation indices (VIs), texture feature indices (TFIs), and canopy temperature (T) were extracted, and a Random Forest (RF) model was employed to identify dominant features. Subsequently, temporal feature sequences were constructed to develop yield prediction models under single-, dual-, and multi-temporal scenarios. Five deep learning models, along with an RF baseline, were evaluated. This study demonstrates a feature-driven deep learning framework effective for small-to-moderate datasets. Our results indicate that the convolutional neural network–long short-term memory (CNN–LSTM) model achieved the best performance, with a coefficient of determination (R 2 ) of 0.87 and root mean square error (RMSE) of 0.55 kg/m 2 under multi-temporal conditions. The integration of VIs, TFIs, and T consistently outperformed single-source features, thereby highlighting the advantages of combining spectral, structural, and thermal information. Among individual growth stages, the heading stage yielded the best single-temporal performance (R 2 = 0.83), whereas the combination of tillering, heading, and grain filling achieved the highest overall accuracy. In this study, we provide an effective framework for UAV-based rice yield prediction and offer practical guidance for selecting optimal observation stages and data combinations, thereby supporting precision agriculture and crop management.
Related
- Winter wheat yield prediction using temporal fusion of multispectral imagery and LiDAR point clouds acquired from a ground-based phenotyping platform
- Spoilage assessment of chicken breast fillets by means of fourier transform infrared spectroscopy and multispectral image analysis
- Multispectral UAV dataset for multi-temporal mangrove monitoring in the Cirebon Coastal Area, Indonesia (2025-2026).
- Multi-temporal UAV multispectral phenotyping and machine learning for modeling cotton defoliation dynamics across diverse accessions
- Fine-scale grassland land-cover classification using UAV RGB and multispectral imagery with staged feature selection
BibTeX
@article{tian2026rice,
title = {{Rice multi-variety yield prediction using multi-temporal fusion of thermal infrared and multispectral UAV data based on random Forest feature selection}},
author = {Xin Tian and Fulang Cen and Junshuo Wei and Quanzhi Zhao and Zhenran Gao},
journal = {Ecological Informatics},
year = {2026},
month = jun,
doi = {10.1016/j.ecoinf.2026.103894},
url = {https://doi.org/10.1016/j.ecoinf.2026.103894},
note = {CC BY},
}
RIS
TY - JOUR
TI - Rice multi-variety yield prediction using multi-temporal fusion of thermal infrared and multispectral UAV data based on random Forest feature selection
AU - Xin Tian
AU - Fulang Cen
AU - Junshuo Wei
AU - Quanzhi Zhao
AU - Zhenran Gao
PY - 2026
DA - 2026/06/20
JO - Ecological Informatics
DO - 10.1016/j.ecoinf.2026.103894
UR - https://doi.org/10.1016/j.ecoinf.2026.103894
AB - Rice is a major staple crop that feeds nearly half of the global population; thus, its accurate yield estimation is critical for food security and agricultural management. To improve prediction accuracy and applicability, in this study, we propose a multi-temporal fusion framework that integrates unmanned aerial vehicle (UAV)-based multispectral and thermal infrared data with feature selection and deep learning models. UAV data were collected from 57 rice varieties at four key growth stages (tillering, heading, grain filling, and maturity). Vegetation indices (VIs), texture feature indices (TFIs), and canopy temperature (T) were extracted, and a Random Forest (RF) model was employed to identify dominant features. Subsequently, temporal feature sequences were constructed to develop yield prediction models under single-, dual-, and multi-temporal scenarios. Five deep learning models, along with an RF baseline, were evaluated. This study demonstrates a feature-driven deep learning framework effective for small-to-moderate datasets. Our results indicate that the convolutional neural network–long short-term memory (CNN–LSTM) model achieved the best performance, with a coefficient of determination (R 2 ) of 0.87 and root mean square error (RMSE) of 0.55 kg/m 2 under multi-temporal conditions. The integration of VIs, TFIs, and T consistently outperformed single-source features, thereby highlighting the advantages of combining spectral, structural, and thermal information. Among individual growth stages, the heading stage yielded the best single-temporal performance (R 2 = 0.83), whereas the combination of tillering, heading, and grain filling achieved the highest overall accuracy. In this study, we provide an effective framework for UAV-based rice yield prediction and offer practical guidance for selecting optimal observation stages and data combinations, thereby supporting precision agriculture and crop management.
SN - 1574-9541
SN - 1878-0512
ER -