Remote Sensing, Artificial Intelligence, and the Internet of Things in Precision Agriculture: A Critical Review of Opportunities, Challenges, and Pathways to Sustainable Intensification
批判性综述遥感、AI与物联网在精准农业中对可持续集约化的实际成效与局限。
关注点该综述指出AI模型在独立田间测试中精度显著下降、小农场收益更小且不稳定,提示评估智慧农业方案须以多季田间联合验证为准。
Hierarchical prediction of SPAD values of flue-cured tobacco at different growth stages under fertigation conditions based on unmanned aerial vehicle multispectral remote sensing data
Hierarchical prediction of SPAD values of flue-cured tobacco at different growth stages under fertigation conditions based on unmanned aerial vehicle multispectral remote sensing data。Smart Agricultural Technology
关注点该研究用无人机多光谱数据按烤烟不同生育期分层反演SPAD值,提示单一模型跨生育期直接套用可能失真,适用于水肥一体化烟田的精准氮素诊断场景。
Mapping soil management zones in green bean (Phaseolus vulgaris L.) fields based on bare soil reflectance and vegetation indices
Mapping soil management zones in green bean (Phaseolus vulgaris L.) fields based on bare soil reflectance and vegetation indices。Precision Agriculture
关注点该文以裸土反射率与植被指数联合划分菜豆田土壤管理分区,其适用性依赖特定田块与传感器条件,推广前需本地验证。
Advancing Plot-level Crop Yield Prediction with Satellite Remote Sensing and Adaptive Temporal Representation Learning
Advancing Plot-level Crop Yield Prediction with Satellite Remote Sensing and Adaptive Temporal Representation Learning。Smart Agricultural Technology
关注点该研究将自适应时序表征学习用于地块尺度卫星遥感估产,关注点在于其方法在田块破碎、多云雾地区的适用边界与验证数据规模。
Mitigating interannual domain shift in UAV-based rice LAI estimation: Annual-reset stage-progressive calibration under in-season information constraints
Mitigating interannual domain shift in UAV-based rice LAI estimation: Annual-reset stage-progressive calibration under in-season information constraints。Computers and Electronics in Agriculture
关注点提出年度重置、分阶段递进的校准策略,在当季信息受限条件下缓解跨年域偏移,适用于多年度无人机水稻长势监测模型的可迁移性维护。
Plant disease detection and recognition using hyperspectral remotely sensed data: A review
综述高光谱遥感用于植物病害检测的数据获取、预处理与建模方法及研究空白。
Integrating UAV LiDAR-Derived Structural and Spectral Data for Improved Aboveground Biomass Estimation in Konjac
Integrating UAV LiDAR-Derived Structural and Spectral Data for Improved Aboveground Biomass Estimation in Konjac。Smart Agricultural Technology
关注点该研究将无人机LiDAR结构特征与光谱数据融合用于魔芋地上生物量估算,其精度增益与适用条件需结合具体品种、生育期和地块验证,不宜直接外推至其他作物。
High-precision cotton leaf area index and boll trait estimation using UAV multiview imaging and 3D point cloud fusion
High-precision cotton leaf area index and boll trait estimation using UAV multiview imaging and 3D point cloud fusion。Computers and Electronics in Agriculture
关注点该研究将无人机多视角成像与三维点云融合用于棉花叶面积指数和棉铃性状估算,其精度提升幅度与适用生育期、品种范围尚需原文数据核实。
Maize Leaf Area Index Retrieval Using MCMC-Calibrated PROSAIL with UAV Sensor Spectral Response Weighting
Leaf area index (LAI) is an important indicator of crop growth and development. However, empirical LAI retrieval models require numerous ground samples, whereas physical models involve complex parameterization. Moreover, representing UAV sensor bands solely by their center wavelengths may cause discrepancies between PROSAIL simulations and multispectral observations. This study developed a growth-stage-specific framework for maize LAI retrieval by integrating UAV imagery, ground measurements, and sensor-response-weighted PROSAIL simulations. First, continuous PROSAIL spectra were convolved with the spectral response functions of the Parrot Sequoia camera, and key parameters were calibrated using Markov chain Monte Carlo (MCMC). Posterior-constrained simulations were then combined with ground measurements to train a random forest model, forming the MCMC-PROSAIL+RF method. Separate retrieval models were developed for the vegetative and reproductive stages. Adding MCMC-calibrated simulations to the ground dataset increased R ² by 0.02 and 0.22 at the two stages, respectively. Compared with center-wavelength reflectance, spectral-response weighting increased R ² from 0.96 to 0.97 at the vegetative stage and from 0.50 to 0.65 at the reproductive stage. At the reproductive stage, RMSE decreased from 0.45 to 0.38 and MAE from 0.33 to 0.32. The final models achieved R ² values of 0.97 and 0.65 at the vegetative and reproductive stages, respectively. These results indicate that combining sensor-response-aware spectral simulation with growth-stage-specific Bayesian parameter calibration improves the consistency between physical simulations and UAV observations and enhances maize LAI retrieval under limited ground sampling.
[预印本] 基于谷歌AlphaEarth嵌入与田块边界图的摩洛哥全国尺度作物类型制图(National-Scale Crop Type Mapping in Morocco Using Google AlphaEarth Foundations Embeddings and the Fields of the World Field Boundary Map)
用谷歌AlphaEarth嵌入与田块边界图实现摩洛哥全国尺度三类作物制图。
Pixel-oriented prediction of soil moisture using graph neural network and remote sensing data- a case study in Karnataka, India
用图神经网络结合遥感多源数据预测印度卡纳塔克邦逐像元土壤湿度。
GIS-Based Agricultural Potential Zoning Using Remote Sensing Indicators and the Analytic Hierarchy Process: A Case Study of Blida Wilaya, Northern Algeria
用GIS与遥感指标结合AHP对阿尔及利亚Blida省做农业潜力分区并验证。
Estimation of Rice Leaf Area Index Coupling Feature Selection from Multi-Temporal UAV Imagery with Machine Learning
融合多时相无人机影像特征选择与机器学习,实现水稻叶面积指数高精度估算。
关注点该研究用RF结合SHAP-RFE特征选择,在全生育期实现水稻LAI估算R²=0.9197、rRMSE=16.85%,可为变量施肥与长势监测提供参考,但结论基于控制试验特定品种与氮梯度,外推需验证。
Field-scale soil salinity mapping from UAV multispectral imagery by disentangling soil and vegetation spectral responses
Field-scale soil salinity mapping from UAV multispectral imagery by disentangling soil and vegetation spectral responses。Smart Agricultural Technology
关注点该研究以无人机多光谱影像分离土壤与植被光谱贡献来反演田块尺度盐渍化,适用边界在于单一田块尺度与特定作物覆盖条件,跨区域推广前需验证。
Data-driven insights into priority predictors associated with conception success in indigenous goats in Uganda
用五种机器学习算法分析乌干达本地山羊900条配种记录,识别与受孕成功相关的优先预测因子。
A Review of Image-Based and AI-Driven Water Level Monitoring: Methods, Datasets, and Deployment Considerations
综述93项基于图像与AI的水位监测研究,梳理方法、数据集与部署问题。
Optimizing remote sensing image classification: a grey relational analysis and three-way decision approach
Optimizing remote sensing image classification: a grey relational analysis and three-way decision approach。International Journal of Machine Learning and Cybernetics
关注点该文将灰色关联分析与三支决策结合用于遥感图像分类优化,属方法层面探索;摘要未给出农业数据集与精度对比,适用性需结合具体作物遥感任务验证。
Fusion of multi-scale Fourier spectrum texture indices and vegetation indices for improved rice aboveground biomass estimation
Fusion of multi-scale Fourier spectrum texture indices and vegetation indices for improved rice aboveground biomass estimation。Computers and Electronics in Agriculture
关注点该研究将多尺度傅里叶频谱纹理指数与植被指数融合用于水稻地上生物量估算,适用边界在于其精度依赖特定试验数据与传感器条件,跨区域推广前需独立验证。
COMPARATIVE EVALUATION OF CLASSICAL MACHINE LEARNING AND DEEP LEARNING FOR CROP TYPE CLASSIFICATION FROM SENTINEL-2 NDVI TIME SERIES
Accurate crop type classification from satellite imagery underpins food security monitoring, subsidy verification, and precision agriculture. The objective of this study is to establish which classification paradigm – classical machine learning or deep learning – provides higher accuracy and better cross-domain transferability for crop type mapping from Sentinel-2 NDVI time series, focusing on the underrepresented conditions of Ukrainian agriculture. Nine classifiers – six classical (Random Forest, SVM, XGBoost, LightGBM, KNN, Gradient Boosting) and three deep-learning-based (Temporal CNN, Bidirectional LSTM, Transformer encoder) – are evaluated on the multi-country AgriSen-COG benchmark and independently tested on a Ukrainian farm dataset (Krynychanska community, Dnipro region; four crops, 2025 season, 38,867 samples). Performance is assessed through overall accuracy (OA), balanced accuracy, weighted/ macro F1, Cohen’s kappa, Matthews correlation coefficient, ROC AUC, and per-class F1. The scientific novelty lies in: (a) the first unified nine-model comparison under identical protocols on a standardized European benchmark with simultaneous cross-domain transfer evaluation on independent Ukrainian field data; (b) quantification of the domain gap between Western-European benchmark and Ukrainian agricultural conditions for both classical and deep learning methods. Results show that deep learning models outperform classical classifiers on the benchmark, with TempCNN achieving the highest OA (0.839) – 3.2 % above the best classical model (XGBoost, 0.813). Upon cross-domain transfer, accuracy degrades substantially across all models (mean ΔOA = -56.4 %), yet deep learning retains a relative advantage: DL models lose on average 52.1 % vs. 58.5 % for classical ML, and TempCNN transfers best (OA = 0.441, +34.9 % relative to the best classical transfer). Per-class analysis reveals that only maize transfers reliably (F1 ≥ 0.82), while rapeseed and sunflower require domain adaptation. Conclusions. The practical significance of the results consists in providing evidence-based guidelines for selecting classification methods in operational crop monitoring systems, demonstrating that TempCNN is the preferred model for both in-domain accuracy and cross-domain robustness, and confirming that domain adaptation is essential before deploying benchmark-trained classifiers in Ukrainian agriculture.
关注点TempCNN在基准上总体精度0.839,跨域迁移后仍最优(0.441),但所有模型平均精度下降56.4%;仅玉米迁移可靠(F1≥0.82),油菜和向日葵需域适应。
Use of Remote Sensing Techniques for Agriculture Survey: A Review
This article reviews the integration of remote sensing technologies into agricultural surveys to address the limitations of conventional data collection, such as high costs, questionable reliability, and lack of timeliness. It examines how electromagnetic radiation, spectral signatures, and various sensor systems (including IRS-1C, LISS-III, PAN, and WiFS) are utilized to monitor dynamic agricultural resources. The review highlights specific applications across three main domains: crop surveys (e.g., crop identification, yield forecasting, and disease infestation), range surveys (e.g., forest delineation and soil moisture), and livestock surveys. The study concludes that remote sensing provides a rapid, cost-effective, and comprehensive synoptic view essential for effective local and national agricultural planning and resource management.