Explainable artificial intelligence has been extensively studied recently. However, the research of interpretable methods in the agricultural field has not been systematically studied. We studied the interpretability of deep learning models in different agricultural classification tasks based on the fruit leaves dataset. The purpose is to explore whether the classification model is more inclined to extract the appearance characteristics of leaves or the texture characteristics of leaf lesions during the feature extraction process. The dataset was arranged into three experiments with different categories. In each experiment, the VGG, GoogLeNet, and ResNet models were used and the ResNet-attention model was applied with three interpretable methods. The results show that the ResNet model has the highest accuracy rate in the three experiments, which are 99.11%, 99.4%, and 99.89%, respectively. It is also found that the attention module could improve the feature extraction of the model, and clarify the focus of the model in different experiments when extracting features. These results will help agricultural practitioners better apply deep learning models to solve more practical problems.

The leaf chlorophyll content (LCC) is a critical index to characterize crop growth conditions, photosynthetic capacity, and physiological status. Its dynamic change characteristics are of great significance for monitoring crop growth conditions and understanding the process of material and energy exchange between crops and the environment. Extensive research has focused on LCC retrieval with hyperspectral data onboard various sensor platforms. Nevertheless, limited attention has been paid to LCC inversion from multispectral data, such as the data from Landsat-8, and the potentials and capabilities of the data for crop LCC estimation have not been fully explored. The present study made use of Landsat-8 Operational Land Imager (OLI) imagery and the corresponding field experimental data to evaluate their capabilities and potentials for LCC modeling using four different retrieval methods: vegetation indices (VIs), machine learning regression algorithms (MLRAs), lookup-table (LUT)-based inversion, and hybrid regression approaches. The results showed that the modified triangular vegetation index (MTVI2) exhibited the best estimate accuracy for LCC retrieval with a root mean square error (RMSE) of 5.99 μg/cm² and a relative RMSE (RRMSE) of 10.49%. Several other vegetation indices that were established from red and near-infrared (NIR) bands also exhibited good accuracy. Models established from Gaussian process regression (GPR) achieved the highest accuracy for LCC retrieval (RMSE = 5.50 μg/cm², RRMSE = 9.62%) compared with other MLRAs. Moreover, red and NIR bands outweighed other bands in terms of GPR modelling. LUT-based inversion methods with the “K(x) = −log (x) + x” cost function that belongs to the “minimum contrast estimates” family showed the best estimation results (RMSE = 8.08 μg/cm², RRMSE = 14.14%), and the addition of multiple solution regularization strategies effectively improved the inversion accuracy. For hybrid regression methods, the use of active learning (AL) techniques together with GPR for LCC modelling significantly increased the estimation accuracy, and the combination of entropy query by bagging (EQB) AL and GPR had the best accuracy for LCC estimation (RMSE = 12.43 μg/cm², RRMSE = 21.77%). Overall, our study suggest that Landsat-8 OLI data are suitable for crop LCC retrieval and could provide a basis for LCC estimation with similar multispectral datasets.

Powdery mildew, caused by the fungus Blumeria graminis, is a major winter wheat disease in China. Accurate delineation of powdery mildew infestations is necessary for site-specific disease management. In this study, high-resolution multispectral imagery of a 25 km² typical outbreak site in Shaanxi, China, taken by a newly-launched satellite, SPOT-6, was analyzed for mapping powdery mildew disease. Two regions with high representation were selected for conducting a field survey of powdery mildew. Three supervised classification methods—artificial neural network, mahalanobis distance, and maximum likelihood classifier—were implemented and compared for their performance on disease detection. The accuracy assessment showed that the ANN has the highest overall accuracy of 89%, following by MD and MLC with overall accuracies of 84% and 79%, respectively. These results indicated that the high-resolution multispectral imagery with proper classification techniques incorporated with the field investigation can be a useful tool for mapping powdery mildew in winter wheat.

The establishment and application of a spectral library is a critical step in the standardization and automation of remote sensing interpretation and mapping. Currently, most spectral libraries are designed to support the classification of land cover types, whereas few are dedicated to agricultural remote sensing monitoring. Here, we gathered spectral observation data on plants in multiple experimental scenarios into a spectral database to investigate methods for crop classification (16 crop species) and status monitoring (tea plant and rice growth). We proposed a set of screening methods for spectral features related to plant classification and status monitoring (band reflectance, vegetation index, spectral differentiation, spectral continuum characteristics) that are based on ISODATA and JM distance. Next, we investigated the performance of different machine learning classifiers in the spectral library application, including K-nearest neighbor (KNN), Random Forest (RF), and a genetic algorithm coupled with a support vector machine (GA-SVM). The optimal combination of spectral features and the classifier with the highest classification accuracy were selected for crop classification and status monitoring scenarios. The GA-SVM classifier performed the best, which produced an accuracy of OAA = 0.94, Kappa = 0.93 for crop classification in a complex scenario (crops mixed with 71 non-crop plant species), and promising accuracies for tea plant growth monitoring (OAA = 0.98, Kappa = 0.97) and rice growth stage monitoring (OAA = 0.92, Kappa = 0.90). Therefore, the establishment of a plant spectral library combined with relevant feature extraction and a classification algorithm effectively supports agricultural monitoring by remote sensing.