Grain protein content (GPC) is a crucial indicator for evaluating rice quality and its commercial value. Establishing a rapid and non-destructive method for estimating rice GPC was established, so as to provide theoretical foundations and technical support for smart breeding and precision crop management.

This study employed a drone equipped with both an RGB camera and a multispectral camera to collect RGB and multispectral imagery, along with ground-measured grain protein content (GPC) data, from the heading to maturity stages of 522 rice breeding material accessions from 2022 to 2023. The Gram-Schmidt image fusion method was applied to process the RGB and multispectral images for generating fused images. Spectral and texture features extracted from the original multispectral images were combined with fused image features, and three machine learning regression algorithms—Random Forest (RF), Extreme Gradient Boosting (XGBoost), and Gradient Boosting Regression (GBR)—were employed to construct GPC estimation models.

The R-band of the RGB images contained richer image information. Vegetation indices derived from the fused R-band exhibited higher correlations with GPC than those calculated from the original multispectral data. The mean texture (Mean) appeared most frequently in texture index construction (accounting for 63.16%), with the MEA₅₆₀-MEA₈₄₀ index showing certain correlations with GPC across different rice types (Huaian conventional japonica: |r²|=0.28; Rugao hybrid japonica: |r²|=0.20). Using a combination of multispectral image features, texture features, and fused image features as input parameters, the GPC estimation models for rice breeding materials achieved higher accuracy at the heading stage (R2 cal=0.64) and maturity stage (R2 cal=0.70) than at the filling stage model (R2 cal=0.53). Incorporating fused image features improved GPC estimation accuracy (ΔR2 cal=0.08-0.26) over using original image features. The interannual model of RF outperformed those of XGBoost and GBR in accuracy (RF: R2 val=0.74, RMSE=0.21%; XGBoost: R2 val=0.58, RMSE=0.23%; GBR: R2 val=0.42, RMSE=0.23%).

The integration of UAV image fusion technique and machine learning methods could effectively enhance the estimation accuracy of the grain protein content (GPC) in rice breeding materials. These findings provided a theoretical reference and practical approaches for the precise estimation of rice quality parameters on a large scale.

The seedling condition of wheat reflects the growth status and health level of the seedlings, which is an important basis for yield prediction and field management. Traditional seedling condition evaluation methods relying on manual expertise have limitations in large-scale field applications, such as low efficiency, strong subjectivity, and difficulty in generalization. This study used UAVs equipped with RGB sensors, combined with ground-measured agronomic parameters, to explore a comprehensive wheat seedling condition assessment method that integrated multimodal remote sensing features, so as to provide a technical pathway for large-scale and cross-regional monitoring.

Multi-site and multi-altitude UAV flight tests were designed in Jiangsu Province to capture UAV imagery and to simultaneously collect agronomic parameters, such as tiller number and canopy coverage. Based on vegetation indices and texture features, crop spectral and structural information was extracted. Feature selection was performed using Information Value (IV) and GINI coefficients. Various machine learning models, including random forest (RF), extreme gradient boosting (XGBoost), and gradient boosting decision tree (GBDT), were developed to assess the classification accuracy of seedling condition. The optimal image resolution was determined based on the local variance coefficient to enhance the stability and cross-regional adaptability of the application.

The enhanced green-red difference index (EXGR) showed the best accuracy in identifying canopy coverage during the seedling stage (Pixel Accuracy (PA)=0.69, Specificity (S)=0.83). The green-red ratio index (GRRI) exhibited a significant correlation with tiller number (R²=0.58, relative root mean square error (rRMSE)=0.28). The Random Forest algorithm, which integrates agronomic parameters and remote sensing features, achieved the highest accuracy in seedling condition grade classification (PA=0.85, R=0.86). Tiller number and texture information (E_energy) contributed the most to the seedling condition grade classification (IV>0.70). A flight altitude of (35±5) m was found to be an important reference for obtaining high-quality UAV data (local variance=0.17).

This study constructed a comprehensive wheat seedling condition assessment framework integrating agronomic parameters and remote sensing features, which demonstrated the feasibility and efficiency of portable UAV RGB imagery combined with machine learning methods for cross-regional seedling condition grade monitoring. This method could provide data support and methodological references for regional-scale dynamic seedling condition assessment, crop management strategy formulation, and food security assurance.