Related Experiment Video
Updated: Jun 13, 2025
08:47
Author Spotlight: UAV Remote Sensing for Efficient Invasive Plant Biomass Estimation
Published on: February 9, 2024
1.3K
High-Throughput Phenotyping of Soybean Biomass: Conventional Trait Estimation and Novel Latent Feature Extraction
Mashiro Okada1, Clément Barras1, Yusuke Toda1
1Graduated School of Agricultural and Life Sciences, The University of Tokyo, Tokyo, Japan.
Plant Phenomics (Washington, D.C.)
|September 10, 2024
Summary
We used drone remote sensing and deep learning to estimate soybean biomass traits, accelerating crop breeding. This approach accurately predicts phenotypes and identifies genetically controlled traits for improved soybean varieties.
Area of Science:
- Agricultural Science
- Plant Breeding
- Remote Sensing Technology
Background:
- High-throughput phenotyping accelerates crop breeding by reducing costs and time.
- Soybean (Glycine max) breeding requires efficient methods for assessing biomass-related traits.
Purpose of the Study:
- To develop deep learning models for estimating soybean biomass traits using UAV remote sensing data.
- To assess the accuracy of these models and explore the utility of extracted latent features in breeding.
Main Methods:
- Utilized convolutional neural networks (CNNs) with UAV-derived RGB images and digital surface models.
- Trained models using manually measured phenotypes of 5 biomass traits (dry weight, stem length, nodes, branches, plant height).
- Assessed model accuracy via 10-fold cross-validation and evaluated latent features using genomic prediction.
Main Results:
- CNN models accurately estimated soybean phenotypes for all 5 biomass-related traits simultaneously.
- Deep learning effectively extracted correlated features from remote sensing data.
- Low-dimensional latent features showed genetic control, indicating potential for breeding applications.
Conclusions:
- UAV remote sensing combined with deep learning offers a powerful tool for high-throughput phenotyping in soybean.
- Extracted latent features hold promise for genomic prediction and accelerating soybean genetic improvement.
- This integrated approach can significantly enhance the efficiency of crop breeding programs.
