Related Experiment Video
Updated: Jun 7, 2025
08:47
Author Spotlight: UAV Remote Sensing for Efficient Invasive Plant Biomass Estimation
Published on: February 9, 2024
1.3K
Estimating Summer Maize Biomass by Integrating UAV Multispectral Imagery with Crop Physiological Parameters
Qi Yin1,2, Xingjiao Yu1,2, Zelong Li1,2
1Key Laboratory of Agricultural Soil and Water Engineering in Arid and Semiarid Areas, Ministry of Education, Northwest AF University, Yangling 712100, China.
Plants (Basel, Switzerland)
|November 9, 2024
Summary
Unmanned aerial vehicle spectral data alone has limited accuracy for predicting summer maize aboveground biomass (AGB). Adding plant height data significantly improves prediction accuracy, with the LSTM model showing the best performance.
Area of Science:
- Agricultural Remote Sensing
- Crop Physiology
- Machine Learning in Agriculture
Background:
- Accurate estimation of summer maize aboveground biomass (AGB) is crucial for agricultural management and yield prediction.
- Traditional AGB measurement methods are inefficient and lack spatial information.
- Unmanned aerial vehicle (UAV) technology offers potential for efficient AGB monitoring, but current prediction accuracy is limited.
Purpose of the Study:
- To develop and evaluate machine learning models for predicting summer maize AGB throughout its growth period using UAV data.
- To assess the contribution of spectral data (color indices, CIS) and plant height (PH) to AGB prediction accuracy.
- To compare the performance of Partial Least Squares Regression (PLSR), Random Forest (RF), and Long Short-Term Memory (LSTM) models.
Main Methods:
- Multispectral images were captured using a DJI Phantom 4 Pro at six key growth stages of summer maize.
- Color indices (CIS) and digital elevation model (DEM) data were extracted from UAV imagery.
- Ground-truth data including AGB and plant height (PH) were collected; PLSR, RF, and LSTM models were trained and validated.
Main Results:
- UAV spectral data (CIS) alone yielded poor AGB prediction accuracy, with LSTM (CIS) achieving the highest R² (0.516-0.649).
- Incorporating plant height (PH) data significantly improved model performance, increasing R² by approximately 25% and reducing RMSE/NRMSE by 20%.
- The LSTM (PH + CIS) model demonstrated the best performance, achieving R² = 0.744, RMSE = 4.833 g, and NRSME = 0.107.
Conclusions:
- Combining UAV-derived spectral data with plant height significantly enhances the prediction accuracy of summer maize aboveground biomass.
- The LSTM model, when integrated with both CIS and PH data, provides a robust and accurate method for AGB monitoring.
- This approach offers a valuable reference for precise, UAV-based remote sensing of crop biomass status.
