Abstract:Photosynthesis is the material basis for crop growth and yield formation. Rapid and real-time monitoring of photosynthetic parameters is essential for crop stress monitoring, light use efficiency assessment, and yield prediction. Methods based on passive optical remote sensing can accurately monitor the photosynthetic traits of crops. However, they are limited in capturing one- or two-dimensional information, and cannot obtain three-dimensional (3D) photosynthetic traits in crops. To address this issue, this study quantifies the 3D photosynthetic trait of wheat based on a digital surface model which fused a light detection and ranging (LiDAR) point cloud with multispectral imagery, both of which were acquired using an unmanned aerial vehicle (UAV). First, we constructed a model to estimate the key photosynthetic parameters (canopy chlorophyll content, fraction of absorbed photosynthetically active radiation, and canopy net photosynthetic rate) based on selected multi-spectral indices in wheat. Second, we applied the model to a point cloud that had been fused with spectral reflectance imagery to generate the 3D spatial and vertical distributions of the above three photosynthetic parameters, and then estimated the 3D photosynthetic traits. Third, we analyzed the 3D photosynthetic traits of wheat in terms of cultivars, growth stages, and cultivation management, and assessed the ability of the derived metrics based on the 3D photosynthetic traits to predict yield. The results showed that the LiDAR point cloud fused with multispectral images could accurately characterize the 3D spatial distribution of the photosynthetic parameters of wheat under different conditions. The photosynthetic traits of the different horizontal layers (upper, middle, and lower) of the canopy were statistically different. In addition, two new developed 3D metrics (CPn_P75th and CCC_P75th), which were derived from the 3D photosynthetic parameters based on the fusion of point cloud and multispectral data, could predict yield with a higher accuracy and greater robustness than traditional methodology of agronomic traits or remote sensing. This study demonstrate that the fusion of LiDAR point cloud and multispectral imagery is feasible for the accurate quantification of 3D photosynthetic traits and the screening of high-yielding genetic cultivars in crops. In addition, we found that multispectral point clouds generated from UAV multispectral imagery using structure from motion (SfM) combined with multi-view stereo (MVS) algorithms (SfM-MVS) can also be used to estimate the 3D spatial distributions of photosynthetic parameters in crops. Despite the fact that the multispectral point cloud by SfM-MVS is sparse and incomplete, this method has potential application in the study of 3D spatial distribution of crop parameters.

Deciphering the contributions of spectral and structural data to wheat yield estimation from proximal sensing

Using Multi-Sensor Data Fusion Techniques and Machine Learning Algorithms for Improving UAV-Based Yield Prediction of Oilseed Rape

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

Identifying the Contributions of Multi-Source Data for Winter Wheat Yield Prediction in China

Improving Wheat Yield Prediction Integrating Proximal Sensing and Weather Data with Machine Learning

Predicting Grain Yield and Protein Content in Wheat by Fusing Multi-Sensor and Multi-Temporal Remote-Sensing Images

Simultaneous Prediction of Wheat Yield and Grain Protein Content Using Multitask Deep Learning from Time-Series Proximal Sensing

Improved Winter Wheat Yield Estimation by Combining Remote Sensing Data, Machine Learning, and Phenological Metrics

A Multimodal and Temporal Network-Based Yield Assessment Method for Different Heat-Tolerant Genotypes of Wheat

Wheat yield estimation using remote sensing data based on machine learning approaches

Optimal Hyperspectral Characteristics Determination For Winter Wheat Yield Prediction

UAV-based multi-sensor data fusion and machine learning algorithm for yield prediction in wheat

Prediction of Field-Scale Wheat Yield Using Machine Learning Method and Multi-Spectral UAV Data

Novel 3D Photosynthetic Traits Derived from the Fusion of UAV LiDAR Point Cloud and Multispectral Imagery in Wheat

Improving grain yield prediction through fusion of multi-temporal spectral features and agronomic trait parameters derived from UAV imagery

Estimating Wheat Yield by Integrating the WheatGrow and PROSAIL Models

Comparison of Attention Mechanism-Based Deep Learning and Transfer Strategies for Wheat Yield Estimation Using Multisource Temporal Drone Imagery

Wheat Yield Prediction Using Machine Learning Method Based on UAV Remote Sensing Data

UAV-Based Hyperspectral and Ensemble Machine Learning for Predicting Yield in Winter Wheat

The Relationship Between Wheat Yield and Sun-Induced Chlorophyll Fluorescence from Continuous Measurements over the Growing Season

Wheat Growth Monitoring and Yield Estimation based on Multi-Rotor Unmanned Aerial Vehicle