UAV-Based Remote Sensing Monitoring of Maize Growth Using Comprehensive Indices
Timely, efficient, and accurate acquisition of crop growth data is crucial for agricultural decision-making and management. This research focuses on maize, utilizing unmanned aerial vehicle (UAV) technology to monitor its growth and gather data on plant height, chlorophyll content, and leaf area ind...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10879513/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Timely, efficient, and accurate acquisition of crop growth data is crucial for agricultural decision-making and management. This research focuses on maize, utilizing unmanned aerial vehicle (UAV) technology to monitor its growth and gather data on plant height, chlorophyll content, and leaf area index. Comprehensive growth monitoring indices, CGMICV and CGMICT, were developed using the coefficient of variation method (CV) and the technique for order preference by similarity to an ideal solution (TOPSIS) based on the coefficient of variation method of empowerment (Coefficient of variation-TOPSIS, CT) respectively. A correlation analysis between twelve vegetation indices and the comprehensive growth monitoring indices was conducted to select the optimal vegetation index for model input. Maize growth inversion models were established using Partial Least Squares Regression (PLSR), Support Vector Regression (SVR), and Random Forest (RF), with model efficacy compared using performance metrics. The models were further interpreted using Shapley additive explanations (SHAP) values. The results indicate that the correlation between the comprehensive growth monitoring indices and vegetation indices was generally higher than that observed with individual indices. Among the three machine learning methods, the RF model demonstrated superior performance, showing the highest accuracy in the growth monitoring model established with CGMICT, with a coefficient of determination (<inline-formula> <tex-math notation="LaTeX">$R^{2}$ </tex-math></inline-formula>) of 0.724, a root mean square error (RMSE) of 0.117, and mean absolute error (MAE) of 0.097. SHAP analysis revealed that GNDVI contributed most significantly to the predictive accuracy of the growth model. Spatial imaging distribution of maize inversion using the optimal CGMICT-RF model revealed discernible differences in overall crop growth. This study confirms the precision and reliability of this approach, providing a valuable reference for maize growth monitoring and regional crop production surveillance. |
|---|---|
| ISSN: | 2169-3536 |