Abstract:Efficient acquisition of crop leaf moisture information holds significant importance for agricultural production. This information provides farmers with accurate data foundations, enabling them to implement timely and effective irrigation management strategies, thereby maximizing crop growth efficiency and yield. In this study, unmanned aerial vehicle (UAV) multispectral technology was employed. Through two consecutive years of field experiments (2021–2022), soybean (Glycine max L.) leaf moisture data and corresponding UAV multispectral images were collected. Vegetation indices, canopy texture features, and randomly extracted texture indices in combination, which exhibited strong correlations with previous studies and crop parameters, were established. By analyzing the correlation between these parameters and soybean leaf moisture, parameters with significantly correlated coefficients (p < 0.05) were selected as input variables for the model (combination 1: vegetation indices; combination 2: texture features; combination 3: randomly extracted texture indices in combination; combination 4: combination of vegetation indices, texture features, and randomly extracted texture indices). Subsequently, extreme learning machine (ELM), extreme gradient boosting (XGBoost), and back propagation neural network (BPNN) were utilized to model the leaf moisture content. The results indicated that most vegetation indices exhibited higher correlation coefficients with soybean leaf moisture compared with texture features, while randomly extracted texture indices could enhance the correlation with soybean leaf moisture to some extent. RDTI, the random combination texture index, showed the highest correlation coefficient with leaf moisture at 0.683, with the texture combination being Variance1 and Correlation5. When combination 4 (combination of vegetation indices, texture features, and randomly extracted texture indices) was utilized as the input and the XGBoost model was employed for soybean leaf moisture monitoring, the highest level was achieved in this study. The coefficient of determination (R2) of the estimation model validation set reached 0.816, with a root-mean-square error (RMSE) of 1.404 and a mean relative error (MRE) of 1.934%. This study provides a foundation for UAV multispectral monitoring of soybean leaf moisture, offering valuable insights for rapid assessment of crop growth.

What problem does this paper attempt to address?

### Problems the paper attempts to solve This paper aims to solve the problem of rapid and accurate monitoring of soybean leaf moisture. Specifically, the research objectives are as follows: 1. **Explore the relationships between soybean leaf moisture, vegetation indices and texture features**: - The study analyzes the soybean leaf moisture data and corresponding multispectral images obtained by UAV multispectral technology, and analyzes the correlations between vegetation indices (VIs) and texture features (TFs) and soybean leaf moisture. - Select parameters with significant correlations (\( p < 0.05 \)) as input variables to build a model. 2. **Develop an efficient method for monitoring soybean leaf moisture**: - Use machine learning algorithms (such as Extreme Learning Machine (ELM), Extreme Gradient Boosting (XGBoost) and Back Propagation Neural Network (BPNN)) to establish a prediction model for soybean leaf moisture content. - By comparing different combinations of input variables (Combination 1: vegetation indices; Combination 2: texture features; Combination 3: randomly extracted texture indices; Combination 4: a combination of vegetation indices, texture features and randomly extracted texture indices), find the optimal model and input variable combination. 3. **Improve the efficiency and accuracy of agricultural irrigation management**: - By providing accurate soybean leaf moisture data, help farmers adjust irrigation strategies in a timely manner, optimize the crop growth environment, and improve yield and water resource utilization efficiency. ### Main research methods 1. **Experimental design**: - The experiment was carried out in 2021 - 2022 at the Institute of Water - saving Agriculture in Arid and Semi - arid Areas, Northwest A&F University, China. - The experiment adopted a split - plot design, including different mulching treatments (straw mulching, ridge film mulching, no mulching) and supplementary irrigation strategies (irrigation at the flowering stage, podding stage, and both the flowering and podding stages). 2. **Data collection and pre - processing**: - Use the DJI Matrice M300 RTK UAV equipped with the MS600 Pro multispectral camera platform to collect multispectral remote sensing data during the soybean flowering period. - Determine the soybean leaf moisture content by the drying method. - Use Yusense Map V2.2.2 software to stitch, geometrically correct and radiometrically pre - process the multispectral images, and then extract spectral reflectance and texture features in ENVI 5.3 software. 3. **Model construction and evaluation**: - Analyze the correlations between vegetation indices, texture features and soybean leaf moisture content, and select significantly correlated parameters as model inputs. - Use three machine learning algorithms, ELM, XGBoost and BPNN, to build models and evaluate model performance through cross - validation. ### Results and conclusions - **Correlation analysis**: - Most vegetation indices have a relatively high correlation coefficient with soybean leaf moisture content, while the contribution of texture features is relatively small. - The randomly combined texture index (RDTI) has the highest correlation coefficient with leaf moisture content, reaching 0.683. - **Model performance**: - When using Combination 4 (a combination of vegetation indices, texture features and randomly extracted texture indices) as input variables and using the XGBoost model for monitoring, the model performance is the best. - The coefficient of determination (\( R^2 \)) of the model validation set reaches 0.816, the root - mean - square error (RMSE) is 1.404, and the mean relative error (MRE) is 1.934%. This study provides a theoretical basis and technical support for UAV multispectral monitoring of soybean leaf moisture, which is helpful for rapidly evaluating crop growth status and optimizing agricultural irrigation management.

Soybean (Glycine max L.) Leaf Moisture Estimation Based on Multisource Unmanned Aerial Vehicle Image Feature Fusion

Monitoring Soybean Soil Moisture Content Based on UAV Multispectral and Thermal-Infrared Remote-Sensing Information Fusion

Unmanned Aerial System-Based Wheat Biomass Estimation Using Multispectral, Structural and Meteorological Data

Mapping Soybean Maturity and Biochemical Traits Using UAV-Based Hyperspectral Images

Toward Multi-Stage Phenotyping of Soybean with Multimodal UAV Sensor Data: A Comparison of Machine Learning Approaches for Leaf Area Index Estimation

Using Fusion of Texture Features and Vegetation Indices from Water Concentration in Rice Crop to UAV Remote Sensing Monitor

Evaluation of Soybean Drought Tolerance Using Multimodal Data from an Unmanned Aerial Vehicle and Machine Learning

Estimation of Soybean Yield by Combining Maturity Group Information and Unmanned Aerial Vehicle Multi-Sensor Data Using Machine Learning

Estimation of soil moisture content under high maize canopy coverage from UAV multimodal data and machine learning

Non-destructive monitoring of maize LAI by fusing UAV spectral and textural features

Improving estimation of maize leaf area index by combining of UAV-based multispectral and thermal infrared data: The potential of new texture index

Improving the accuracy of cotton seedling emergence rate estimation by fusing UAV-based multispectral vegetation indices

Inversion of Soybean Net Photosynthetic Rate Based on UAV Multi-Source Remote Sensing and Machine Learning

Estimating leaf area index using unmanned aerial vehicle data: shallow vs. deep machine learning algorithms.

Estimation of soybean yield parameters under lodging conditions using RGB information from unmanned aerial vehicles

Combining Texture, Color, And Vegetation Indices From Fixed-Wing Uas Imagery To Estimate Wheat Growth Parameters Using Multivariate Regression Methods

Pretrained Deep Learning Networks and Multispectral Imagery Enhance Maize LCC, FVC, and Maturity Estimation

Evaluating soil moisture content under maize coverage using UAV multimodal data by machine learning algorithms

Non-Destructive Monitoring of Peanut Leaf Area Index by Combing UAV Spectral and Textural Characteristics

Crop Monitoring Using Satellite/UAV Data Fusion and Machine Learning

Combining Spectral and Texture Features of UAV Images for the Remote Estimation of Rice LAI throughout the Entire Growing Season