Yan et al. (2026) A camera calibration method driven by unmanned aerial vehicles and high-precision estimation of maize SPAD values under adverse conditions

Identification

• Journal: Frontiers in Plant Science
• Year: 2026
• Date: 2026-09-07
• Authors: Jianwen Yan, Shilong Miao, Xianyue Li, Haibin Shi, Shijie Ding, Zhen Li, Ning Wang
• DOI: 10.3389/fpls.2026.1922896

Research Groups

• National Key Laboratory of Ecological Environment for Soil and Water Engineering in Arid Areas, Inner Mongolia Agricultural University, Hohhot, China
• Inner Mongolia Autonomous Region Engineering Research Center for High-Efficiency Water-Saving Technology Equipment and Water and Soil Environmental Effects, Hohhot, China
• Inner Mongolia Autonomous Region Water Resources Research Institute, Hohhot, China

Short Summary

This study developed a UAV–smartphone cross-device calibration framework for SPAD estimation using UAV multispectral imagery, smartphone RGB images, and ground-measured SPAD data. The random forest model achieved the best UAV-based performance with validation R² values of 0.81, 0.79, and 0.75 across three growth stages.

Objective

• To develop a low-cost and accurate approach for field-scale maize SPAD monitoring and precision water–nitrogen management using cross-device feature calibration between UAV multispectral sensors and smartphone cameras.
• To establish a high-accuracy mapping between heterogeneous sensors and eliminate spectral response differences under arid stress conditions.

Study Configuration

• Spatial Scale: Field scale, with 27 plots measuring 10 m × 4 m arranged in a completely randomized design.
• Temporal Scale: The study was conducted over three growth stages: jointing (V6), tasseling (VT), and grain-filling (R3).

Methodology and Data

• Models used: Regularized least squares, ridge regression, support vector machine, and random forest models were compared for maize SPAD estimation.
• Data sources: UAV multispectral imagery, smartphone RGB images, and ground-measured SPAD data collected at the jointing, tasseling, and grain-filling stages.

Main Results

• The random forest model using combined spectral and texture features achieved the best UAV-based performance with validation R² values of 0.81, 0.79, and 0.75 across the three growth stages.
• After cross-device calibration, smartphone-based SPAD estimation achieved R² values of 0.80, 0.86, and 0.82, respectively.

Contributions

• This study provides a low-cost and accurate approach for field-scale maize SPAD monitoring and precision water–nitrogen management using cross-device feature calibration between UAV multispectral sensors and smartphone cameras.
• The results demonstrate the effectiveness of cross-device calibration in bridging UAV multispectral and smartphone RGB observations under arid stress conditions.

Funding

• This study was funded by the National Key Laboratory of Ecological Environment for Soil and Water Engineering in Arid Areas, Inner Mongolia Agricultural University.

Citation

@article{Yan2026camera,
  author = {Yan, Jianwen and Miao, Shilong and Li, Xianyue and Shi, Haibin and Ding, Shijie and Li, Zhen and Wang, Ning},
  title = {A camera calibration method driven by unmanned aerial vehicles and high-precision estimation of maize SPAD values under adverse conditions},
  journal = {Frontiers in Plant Science},
  year = {2026},
  doi = {10.3389/fpls.2026.1922896},
  url = {https://doi.org/10.3389/fpls.2026.1922896}
}