[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"item-3354":3,"related-3354":87},{"id":4,"title":5,"url":6,"summary":7,"summary_zh":8,"content":9,"source_name":10,"source_url":6,"published_at":11,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":15,"score_detail":16,"sources":23,"tags":25,"search_phrases":31,"slug":34,"view_count":35,"doi":36,"paper":37,"created_at":86},3354,"Optimizing canopy nitrogen retrieval in wheat from high-resolution UAV hyperspectral data using RTM-based Gaussian process regression","https:\u002F\u002Fdoi.org\u002F10.1080\u002F01904167.2026.2736627","The canopy nitrogen content (CNC) is one of the vital crop health indicators directly influencing the crop growth, yield, and nutritional status. Unmanned aerial vehicles (UAVs) that acquired hyperspectral imagery opened a new path for nondestructive, accurate, and timely estimation of CNC, facilitating the adoption of smart agro practices in precision agriculture for site-specific nutrient management. With the purpose of processing such hyperspectral data into CNC maps, two hybrid retrieval approaches, protein-based (CNCprot) and chlorophyll-based (CNCchl) models, were built by combining a radiative transfer model with a machine learning Gaussian process regression (GPR) algorithm. The GPR retrieval models were trained using synthetic spectral data from the PROSAIL-PRO model, and the sampling was optimized using the Euclidean distance-based diversity (EBD) active learning (AL) technique. The CNC models were applied to a UAV-borne hyperspectral image in the spectral range of 400-1000 nm with an ultrahigh spatial resolution of 4 cm, acquired over the experimental wheat field of Indian Council of Agricultural Research (ICAR) -Indian Agricultural Research Institute (IARI), New Delhi, India. The GPR models yielded good prediction accuracies with R2 values of 0.74 and 0.56 and NRMSE of 13.62 and 23.82% for CNCchl and CNCprot, respectively. The consistent prediction accuracies and low associated uncertainties revealed that CNC was more accurately mapped using chlorophyll-based CNC retrieval applied to a Vis-NIR-based hyperspectral data. These promising results demonstrated the possibility of applying hybrid GPR models at ultrahigh resolution and the operational viability of UAV datasets for regular CNC monitoring.","冠层氮含量（CNC）是直接影响作物生长、产量和营养状况的重要作物健康指标之一。无人机（UAV）获取的高光谱影像为无损、精准、及时地估算CNC开辟了新途径，有助于在精准农业中采用智能农业实践进行定点养分管理。为了将此类高光谱数据处理为CNC分布图，本研究通过将辐射传输模型与机器学习高斯过程回归（GPR）算法相结合，构建了两种混合反演方法：基于蛋白质的模型（CNCprot）和基于叶绿素的模型（CNCchl）。GPR反演模型使用PROSAIL-PRO模型生成的合成光谱数据进行训练，并采用基于欧氏距离的多样性（EBD）主动学习（AL）技术优化采样。CNC模型被应用于一幅无人机载高光谱影像，光谱范围为400-1000 nm，空间分辨率高达4 cm，该影像获取自印度农业研究理事会（ICAR）-印度农业研究所（IARI）位于印度新德里的实验麦田。GPR模型取得了良好的预测精度，CNCchl和CNCprot的R²值分别为0.74和0.56，NRMSE分别为13.62%和23.82%。一致的预测精度和较低的相关不确定性表明，将基于叶绿素的CNC反演应用于基于可见光-近红外（Vis-NIR）的高光谱数据能够更准确地制图CNC。这些有前景的结果证明了在超高分辨率下应用混合GPR模型的可能性，以及无人机数据集用于常规CNC监测的可行性。",null,"Journal of Plant Nutrition","2026-09-23T00:00:00Z","论文",10,false,79,{"impact":17,"substance":18,"depth":17,"authority":19,"freshness":20,"relevant":21,"comment":22},18,22,13,8,1,"将辐射传输模型与高斯过程回归结合，用4厘米超高分辨率无人机高光谱实现小麦冠层氮素精准反演，方法新颖、结论可靠，对精准农业养分管理有实操参考价值。",[24],{"name":10,"url":6},[26,27,28,29,30],"智慧农业","无人机","小麦","遥感","氮素监测",[32,33],"无人机 高光谱 小麦 氮素","PROSAIL 高斯过程回归 小麦","无人机高光谱小麦氮素-3354",0,"10.1080\u002F01904167.2026.2736627",{"doi":36,"openalex_id":38,"authors":39,"venue":10,"cited_by_count":35,"oa_url":9,"card":78,"direction":84,"ingested_from":85},"W7214104831",[40,42,44,47,50,52,54,57,60,63,65,68,70,72,75],{"name":41,"orcid":9},"R. G. Rejith",{"name":43,"orcid":9},"Rabi N. Sahoo",{"name":45,"orcid":46},"Shalini Gakhar","https:\u002F\u002Forcid.org\u002F0000-0001-5717-1714",{"name":48,"orcid":49},"Jochem Verrelst","https:\u002F\u002Forcid.org\u002F0000-0002-6313-2081",{"name":51,"orcid":9},"Amrita Bhandari",{"name":53,"orcid":9},"Tarun Kondraju",{"name":55,"orcid":56},"Rajeev Ranjan","https:\u002F\u002Forcid.org\u002F0000-0003-2233-9147",{"name":58,"orcid":59},"Mahesh C. Meena","https:\u002F\u002Forcid.org\u002F0000-0001-5386-883X",{"name":61,"orcid":62},"Abir Dey","https:\u002F\u002Forcid.org\u002F0000-0002-5009-9518",{"name":64,"orcid":9},"Joydeep Mukherjee",{"name":66,"orcid":67},"Sudhir Kumar","https:\u002F\u002Forcid.org\u002F0000-0002-1089-7435",{"name":69,"orcid":9},"Mahesh Kumar",{"name":71,"orcid":9},"Raju Dhandapani",{"name":73,"orcid":74},"Anchal Dass","https:\u002F\u002Forcid.org\u002F0000-0003-3909-1803",{"name":76,"orcid":77},"Viswanathan Chinnusamy","https:\u002F\u002Forcid.org\u002F0000-0003-2174-9064",{"tldr":79,"method":80,"finding":81,"direction":82,"opportunity":83},"用PROSAIL-PRO与高斯过程回归从无人机高光谱数据反演小麦冠层氮含量。","PROSAIL-PRO合成光谱训练GPR，EBD主动学习采样，4cm无人机高光谱","叶绿素基模型精度更高，R²达0.74，NRMSE为13.62%，优于蛋白基模型。","农业遥感与作物表型","可探索多生育期、多品种迁移能力及主动学习采样策略对反演精度的提升空间。","智慧农业 \u002F 农业物联网","openalex","2026-09-24T23:30:10.311254Z",{"total":88,"page":21,"page_size":88,"items":89},6,[90,137,166,196,226,270],{"id":91,"title":92,"url":93,"summary":94,"summary_zh":95,"content":9,"source_name":96,"source_url":93,"published_at":97,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":98,"score_detail":99,"sources":103,"tags":105,"search_phrases":107,"slug":110,"view_count":35,"doi":111,"paper":112,"created_at":136},1644,"DMP-UNet: a multimodal fusion network for high-precision nitrogen status monitoring in wheat from UAV multispectral images","https:\u002F\u002Fdoi.org\u002F10.3389\u002Ffpls.2026.1909856","Achieving high-precision monitoring of wheat nitrogen levels is essential for boosting yield and grain quality, yet the remote sensing community has largely confined itself to single-modal data sources, missing the synergistic benefits of multimodal fusion. To address this, we propose a multimodal feature fusion network that simultaneously utilizes RGB, multispectral, and vegetation index data. The network employs dual-branch encoders to extract modality‑specific features from fused images composed of NDVI, green, and near‑infrared bands; a deformable bidirectional cross‑attention module aligns pixel‑level features and enhances cross‑modal interactions; a modality‑aware pyramid attention feature pyramid network fuses semantic information across scales and modalities; and a dual‑path decoder separates background from wheat regions, with a foreground decoder guided by vegetation attention maps to classify nitrogen status into four levels: Enrichment, Optimum Level, Moderate Nitrogen Deficiency, and Severe Nitrogen Deficiency. Experimental results demonstrate that our method outperforms state‑of‑the‑art models such as DeepLabV3+, U‑Net, U‑Net++, and Attention UNet. Modality contribution analysis confirms the complementary roles of RGB and MS, while ablation studies validate the effectiveness of each key module. These findings confirm the practicality and efficiency of the proposed approach for precise nitrogen monitoring in wheat, highlighting its potential for optimizing fertilization strategies in precision agriculture.","实现对小麦氮素水平的高精度监测对于提高产量和籽粒品质至关重要，然而遥感领域长期局限于单一模态数据源，未能充分利用多模态融合的协同优势。为此，我们提出了一种多模态特征融合网络，同时利用RGB、多光谱和植被指数数据。该网络采用双分支编码器，从由NDVI、绿光和近红外波段组成的融合图像中提取模态特定特征；可变形双向交叉注意力模块用于对齐像素级特征并增强跨模态交互；模态感知金字塔注意力特征金字塔网络跨尺度和跨模态融合语义信息；双路径解码器将背景与小麦区域分离，其中前景解码器由植被注意力图引导，将氮素状态分为四个等级：富集、最适水平、中度缺氮和重度缺氮。实验结果表明，我们的方法优于DeepLabV3+、U-Net、U-Net++和Attention UNet等最先进模型。模态贡献分析证实了RGB和多光谱的互补作用，消融研究验证了各关键模块的有效性。这些发现证实了所提方法在小麦精确氮素监测中的实用性和高效性，凸显了其在精准农业中优化施肥策略的潜力。","Frontiers in Plant Science","2026-09-03T00:00:00Z",76,{"impact":17,"substance":100,"depth":17,"authority":19,"freshness":101,"relevant":21,"comment":102},20,7,"提出多模态融合网络提升小麦氮素监测精度，方法新颖，实验验证充分，对精准农业有实际价值。",[104],{"name":96,"url":93},[26,28,106,29,30],"多模态融合",[108,109],"多模态融合 智慧农业 氮素监测 小麦","多模态融合 智慧农业","多模态融合智慧农业氮素监测小麦-1644","10.3389\u002Ffpls.2026.1909856",{"doi":111,"openalex_id":113,"authors":114,"venue":96,"cited_by_count":35,"oa_url":93,"card":131,"direction":82,"ingested_from":85},"W7207858118",[115,117,119,121,123,126,129],{"name":116,"orcid":9},"Qingqing Hong",{"name":118,"orcid":9},"Siqi Cao",{"name":120,"orcid":9},"Bohan Hu",{"name":122,"orcid":9},"Changwei Tan",{"name":124,"orcid":125},"Zhenghua Zhang","https:\u002F\u002Forcid.org\u002F0000-0003-0880-0240",{"name":127,"orcid":128},"Bin Li","https:\u002F\u002Forcid.org\u002F0009-0000-0898-7864",{"name":130,"orcid":9},"Hongwei Zhang",{"tldr":132,"method":133,"finding":134,"direction":82,"opportunity":135},"提出多模态融合网络DMP-UNet，利用无人机多光谱图像高精度监测小麦氮素水平。","双分支编码器、可变形双向交叉注意力、模态感知金字塔注意力FPN、双路径解码器。","该方法优于DeepLabV3+等模型，RGB与多光谱互补，各模块有效。","可探索多模态融合在作物病害、产量预测等领域的应用，或引入时间序列数据提升动态监测能力。","2026-09-04T23:30:33.457309Z",{"id":138,"title":139,"url":140,"summary":141,"summary_zh":9,"content":9,"source_name":142,"source_url":9,"published_at":143,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":144,"score_detail":145,"sources":149,"tags":151,"search_phrases":153,"slug":156,"view_count":35,"doi":9,"paper":157,"created_at":165},1367,"基于无人机多源遥感特征的茶树生长与氮素状况深度学习监测","https:\u002F\u002Fwww.mdpi.com\u002F2072-4292\u002F18\u002F17\u002F2923","南京农业大学园艺学院团队提出融合光谱、纹理与谐波特征的多域融合指标，结合无人机数据与深度学习模型，用于茶树生物量与氮素积累量估算。实验表明多域融合指数显著优于传统光谱指标，可实现对茶园生物量与氮素状况的高精度反演，为近实时茶园监测提供工具。","Remote Sensing · MDPI · 2026-09-01","2026-08-31T16:00:00Z",72,{"impact":146,"substance":100,"depth":17,"authority":147,"freshness":101,"relevant":21,"comment":148},15,12,"研究提出多域融合指标，提升茶树氮素监测精度，对智慧茶园有实用价值。",[150],{"name":142,"url":140},[26,27,29,30,152],"茶树",[154,155],"智慧农业 氮素监测 无人机 茶树","智慧农业 氮素监测","智慧农业氮素监测无人机茶树-1367",{"doi":9,"openalex_id":9,"authors":158,"venue":9,"cited_by_count":35,"oa_url":9,"card":159,"direction":82,"ingested_from":164},[],{"tldr":160,"method":161,"finding":162,"direction":82,"opportunity":163},"融合无人机多源遥感特征与深度学习，高精度估算茶树生物量与氮素积累量。","融合光谱、纹理、谐波特征的多域融合指标，结合无人机数据与深度学习模型。","多域融合指数显著优于传统光谱指标，实现茶园生物量与氮素高精度反演。","可探索多域融合特征在其他作物或不同生长阶段的适用性，或结合时序数据预测氮素动态。","agent","2026-09-02T00:05:07.021728Z",{"id":167,"title":168,"url":169,"summary":170,"summary_zh":9,"content":9,"source_name":171,"source_url":9,"published_at":172,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":173,"score_detail":174,"sources":177,"tags":179,"search_phrases":183,"slug":186,"view_count":35,"doi":9,"paper":187,"created_at":195},3324,"Diag-STFN：全球收获前作物产量预测的诊断时空多模态融合网络——覆盖38国玉米29国小麦（Ecological Informatics 2026）","https:\u002F\u002Fm2.mtmt.hu\u002Fapi\u002Fpublication\u002F37354974?&&labelLang=hun","《Ecological Informatics》2026年第96期：Zhuang等提出Diag-STFN——一种诊断时空多模态融合网络，用于全球收获前作物产量预测。该网络基于数据集特征选择模型结构，以确定是否需要时间趋势耦合和空间模块激活。在三种前置期（早、中、晚季）下，基于覆盖38国玉米和29国小麦的CY-Bench基准数据集进行评估。结果表明，所提方法在所有前置期均实现了两种作物的最低汇总NRMSE，并在MAPE和KGE等补充指标上保持领先。消融研究表明诊断模块选择提供了主要的性能提升；方差分解显示性能差异在国家之间大于模型之间。","《Ecological Informatics》96 (2026) 103860","2026-09-17T00:00:00Z",78,{"impact":17,"substance":18,"depth":17,"authority":175,"freshness":88,"relevant":21,"comment":176},14,"方法新颖、覆盖38国玉米与29国小麦的全球收获前产量预测研究，学术价值突出但产业落地尚早，适合作为前沿技术资讯收录。",[178],{"name":171,"url":169},[26,180,181,28,182,29],"农业人工智能","产量预测","玉米",[184,185],"Diag-STFN 作物产量预测","CY-Bench 玉米 小麦","Diag-STFN作物产量预测-3324",{"doi":9,"openalex_id":9,"authors":188,"venue":9,"cited_by_count":35,"oa_url":9,"card":189,"direction":193,"ingested_from":164},[],{"tldr":190,"method":191,"finding":192,"direction":193,"opportunity":194},"提出诊断式时空多模态融合网络Diag-STFN，实现全球收获前玉米小麦产量预测。","基于CY-Bench基准，按数据特征诊断选择时间趋势与空间模块，覆盖38国玉米2","各前置期均取得最低NRMSE，诊断模块选择贡献最大，国家间差异大于模型间差异。","农业人工智能与决策模型","可探索自适应诊断机制迁移至其他作物，并针对国家间差异开展区域化建模与不确定性量化。","2026-09-24T00:04:02.684732Z",{"id":197,"title":198,"url":199,"summary":200,"summary_zh":9,"content":9,"source_name":201,"source_url":9,"published_at":202,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":203,"score_detail":204,"sources":209,"tags":211,"search_phrases":214,"slug":217,"view_count":35,"doi":9,"paper":218,"created_at":225},2855,"UAV无人机高光谱图像土壤盐度制图(湿度校正)——MDPI Agronomy","https:\u002F\u002Fwww.mdpi.com\u002F2073-4395\u002F16\u002F18\u002F1812","研究评估了6种光谱变换方法(原始反射率Ref、一阶导数FDR、PDS、OSC、FDR+PDS、FDR+OSC),结合3种机器学习算法(KNN、SVR、MLP)。进一步开发了集成这些基础学习者的Stacking集成模型,以提高湿度干扰下土壤盐度反演的精度。结果表明,Stacking模型在评估模型中达到最高的精度和稳定性。FDR+OSC-Stacking组合实现最佳验证性能,R²p=0.87,RMSEP=0.67 mS·cm⁻¹,RPD=2.93。FDR+OSC-Stacking组合成功应用于UAV高光谱图像,用于EC1:5的空间制图。来自吉林大学。","MDPI Agronomy","2026-09-15T00:00:00Z",75,{"impact":205,"substance":206,"depth":207,"authority":19,"freshness":20,"relevant":21,"comment":208},16,21,17,"方法组合新颖、验证指标扎实的无人机高光谱盐分制图研究，属细分领域实质进展，值得精选。",[210],{"name":201,"url":199},[26,27,212,29,213],"机器学习","土壤盐渍化",[215,216],"土壤盐渍化 智慧农业 机器学习 无人机","土壤盐渍化 智慧农业","土壤盐渍化智慧农业机器学习无人机-2855",{"doi":9,"openalex_id":9,"authors":219,"venue":9,"cited_by_count":35,"oa_url":9,"card":220,"direction":82,"ingested_from":164},[],{"tldr":221,"method":222,"finding":223,"direction":82,"opportunity":224},"用无人机高光谱结合Stacking集成模型实现湿度干扰下的土壤盐度制图。","6种光谱变换与KNN、SVR、MLP及Stacking集成，基于UAV高光谱数据","FDR+OSC-Stacking最优，R²p=0.87、RMSEP=0.67 mS·cm⁻¹、RPD","可探索多时相\u002F多传感器融合与迁移学习，提升不同湿度与区域下盐度反演泛化性。","2026-09-18T00:03:30.822732Z",{"id":227,"title":228,"url":229,"summary":230,"summary_zh":231,"content":9,"source_name":232,"source_url":229,"published_at":202,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":233,"score_detail":234,"sources":236,"tags":238,"search_phrases":240,"slug":243,"view_count":35,"doi":244,"paper":245,"created_at":269},2666,"Automated Machine Learning-Driven UAV Remote Sensing for Accurate Winter Wheat Water Content Prediction","https:\u002F\u002Fdoi.org\u002F10.3390\u002Frs18183161","Crop water content is a critical indicator of crop growth status, and its efficient and accurate monitoring is essential for agricultural water resource management. Conventional methods for monitoring winter wheat water content, however, rely mainly on destructive sampling and are labor-intensive and time-consuming. To address these limitations, this study explored the potential of unmanned aerial vehicle (UAV) remote sensing for the rapid and accurate assessment of winter wheat water content. High-resolution canopy remote sensing images were acquired using UAVs equipped with multispectral (MS), RGB, and thermal infrared (TIR) cameras during the flowering and filling stages under six irrigation treatments. Ground-truth sampling data were integrated with the UAV-derived remote sensing data, and an automated machine learning (AutoML) framework—which automatically searches over a range of candidate algorithms and hyperparameters to select the optimal model—was employed to establish regression models for predicting winter wheat moisture content (MC). All models were evaluated using five-fold cross-validation. The results demonstrated that MC prediction performed best during the filling stage, with the TIR sensor achieving the highest accuracy (R2 = 0.812, MAE = 0.0204, RMSE = 0.0274). Compared with single-sensor approaches, multi-sensor fusion further improved predictive performance, achieving an R2 of 0.876, an MAE of 0.0191, and an RMSE of 0.0259 for MC prediction. These findings indicate that UAV-based multi-sensor remote sensing provides an effective means of monitoring winter wheat water content, facilitating timely assessment of crop growth status and optimized irrigation management. Moreover, the use of AutoML enables high-accuracy prediction with minimal human intervention, enhancing the precision of crop water monitoring and advancing precision agriculture.","作物含水量是反映作物生长状况的关键指标，对其进行高效、准确的监测对农业水资源管理至关重要。然而，传统冬小麦含水量监测方法主要依赖破坏性采样，费时费力。为解决这些局限，本研究探索了无人机（UAV）遥感在快速准确评估冬小麦含水量方面的潜力。在六种灌溉处理下，利用搭载多光谱（MS）、RGB和热红外（TIR）相机的无人机在开花期和灌浆期获取了高分辨率冠层遥感图像。将地面实测采样数据与无人机遥感数据相结合，采用自动化机器学习（AutoML）框架——该框架可在一系列候选算法和超参数中自动搜索以选择最优模型——建立预测冬小麦含水量（MC）的回归模型。所有模型均采用五折交叉验证进行评估。结果表明，灌浆期MC预测表现最佳，其中TIR传感器精度最高（R2 = 0.812，MAE = 0.0204，RMSE = 0.0274）。与单传感器方法相比，多传感器融合进一步提升了预测性能，MC预测的R2达到0.876，MAE为0.0191，RMSE为0.0259。这些发现表明，基于无人机的多传感器遥感为监测冬小麦含水量提供了有效手段，有助于及时评估作物生长状况并优化灌溉管理。此外，AutoML的使用使得在最少人工干预下实现高精度预测成为可能，提升了作物水分监测的精度，推动了精准农业发展。","Remote Sensing",80,{"impact":17,"substance":18,"depth":17,"authority":175,"freshness":20,"relevant":21,"comment":235},"AutoML结合无人机多传感器遥感预测冬小麦含水量，方法新颖、数据扎实，对精准灌溉有实用价值，值得进入每日精选。",[237],{"name":232,"url":229},[26,180,28,29,239],"精准灌溉",[241,242],"农业人工智能 智慧农业 精准灌溉 小麦","农业人工智能 智慧农业","农业人工智能智慧农业精准灌溉小麦-2666","10.3390\u002Frs18183161",{"doi":244,"openalex_id":246,"authors":247,"venue":232,"cited_by_count":35,"oa_url":229,"card":264,"direction":82,"ingested_from":85},"W7213246708",[248,251,253,255,258,261],{"name":249,"orcid":250},"Fan Ding","https:\u002F\u002Forcid.org\u002F0000-0001-5482-8290",{"name":252,"orcid":9},"Qian Cheng",{"name":254,"orcid":9},"Fuyi Duan",{"name":256,"orcid":257},"Shuaipeng Fei","https:\u002F\u002Forcid.org\u002F0000-0002-8774-7929",{"name":259,"orcid":260},"Junjie Feng","https:\u002F\u002Forcid.org\u002F0000-0001-8900-2691",{"name":262,"orcid":263},"Zhen Chen","https:\u002F\u002Forcid.org\u002F0000-0002-2847-0042",{"tldr":265,"method":266,"finding":267,"direction":82,"opportunity":268},"用无人机多光谱、RGB和热红外遥感结合AutoML预测冬小麦含水量。","无人机多传感器影像与地面采样，AutoML自动选模型，五折交叉验证。","灌浆期热红外精度最高R²=0.812，多传感器融合提升至R²=0.876。","可探索AutoML与多时相\u002F多源卫星遥感融合，实现区域尺度作物水分精准监测。","2026-09-16T23:30:29.163192Z",{"id":271,"title":272,"url":273,"summary":274,"summary_zh":275,"content":9,"source_name":276,"source_url":273,"published_at":277,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":278,"score_detail":279,"sources":282,"tags":284,"search_phrases":286,"slug":289,"view_count":35,"doi":290,"paper":291,"created_at":319},2500,"Estimation of crop canopy nitrogen content using deep transfer learning with PROSAIL-PRO model and UAV hyperspectral imagery","https:\u002F\u002Fdoi.org\u002F10.1016\u002Fj.compag.2026.112385","Canopy nitrogen content (CNC) is a key indicator for assessing crop nitrogen status. The non-destructive and rapid estimation of CNC can be performed using unmanned aerial vehicle (UAV)-based hyperspectral remote sensing on large-scale; however, it is often constrained by the scarcity of ground truth data relative to model complexity, which can lead to overfitting and poor generalizability. To address these challenges, a novel inversion framework was proposed by integrating the PROSAIL-PRO physical model with deep transfer learning. Generalized physical features were extracted using pre-trained deep neural network on large-scale PROSAIL-PRO simulated spectra. Subsequently, a fine-tuning strategy was applied in which shallow-level parameters were frozen while deep-level parameters were updated, thereby effectively decoupling intrinsic physical features from site-specific variations. Four deep learning architectures including convolutional neural network (CNN), Residual Neural Network (ResNet18), and their respective channel attention-enhanced variants were evaluated and compared with six classical statistical models including a backpropagation neural network (BPNN). The results demonstrated that transfer-learning-optimized deep models significantly outperformed both baseline and traditional methods in terms of accuracy and generalizability. Specifically, the CNN-based transfer model achieved the best performance on the wheat dataset ( R 2 = 0.8597 and RMSECV = 1.7642), whereas the ResNet18-based transfer model achieved the best performance on the maize dataset ( R 2 = 0.6921 and RMSECV = 2.7944). These findings confirmed that PROSAIL-PRO driven deep transfer learning effectively mitigated overfitting in small-sample hyperspectral datasets. By ensuring high-precision inversion and robust generalization, the proposed physics-guided data-driven fusion approach provides a promising solution for crop nitrogen monitoring.","冠层氮含量（Canopy Nitrogen Content, CNC）是评估作物氮素状况的关键指标。基于无人机（UAV）高光谱遥感可以在大尺度上对CNC进行无损快速估测；然而，该任务常受限于地面真值数据相对模型复杂度而言较为稀缺的问题，这会导致过拟合和泛化能力较差。为应对这些挑战，该研究提出了一种将PROSAIL-PRO物理模型与深度迁移学习相结合的新型反演框架。利用在大规模PROSAIL-PRO模拟光谱上预训练的深度神经网络提取广义物理特征。随后，采用冻结浅层参数、更新深层参数的微调策略，从而有效解耦内在物理特征与站点特异性变异。研究评估了四种深度学习架构，包括卷积神经网络（CNN）、残差神经网络（ResNet18）及其各自的通道注意力增强变体，并与包括反向传播神经网络（BPNN）在内的六种经典统计模型进行了比较。结果表明，经迁移学习优化的深度模型在精度和泛化能力方面均显著优于基线和传统方法。具体而言，基于CNN的迁移模型在小麦数据集上取得了最佳性能（R² = 0.8597，RMSECV = 1.7642），而基于ResNet18的迁移模型在玉米数据集上取得了最佳性能（R² = 0.6921，RMSECV = 2.7944）。这些发现证实，PROSAIL-PRO驱动的深度迁移学习有效缓解了小样本高光谱数据集中的过拟合问题。通过确保高精度反演和稳健泛化，所提出的物理引导的数据驱动融合方法为作物氮素监测提供了一种有前景的解决方案。","Computers and Electronics in Agriculture","2026-09-14T00:00:00Z",81,{"impact":17,"substance":18,"depth":17,"authority":175,"freshness":280,"relevant":21,"comment":281},9,"将PROSAIL-PRO物理模型与深度迁移学习结合，缓解小样本高光谱反演过拟合，方法新颖且验证扎实，对作物氮素遥感监测有实用价值。",[283],{"name":276,"url":273},[26,27,180,29,285],"作物氮素监测",[287,288],"作物氮素监测 农业人工智能 智慧农业 无人机","作物氮素监测 农业人工智能","作物氮素监测农业人工智能智慧农业无人机-2500","10.1016\u002Fj.compag.2026.112385",{"doi":290,"openalex_id":292,"authors":293,"venue":276,"cited_by_count":35,"oa_url":273,"card":314,"direction":82,"ingested_from":85},"W7212964562",[294,296,298,300,302,305,307,309,312],{"name":295,"orcid":9},"Jing Zhao",{"name":297,"orcid":9},"Hong Li",{"name":299,"orcid":9},"Junping Liu",{"name":301,"orcid":9},"Wei Chen",{"name":303,"orcid":304},"Xin Guo","https:\u002F\u002Forcid.org\u002F0000-0002-1583-5364",{"name":306,"orcid":9},"Yunlong Wu",{"name":308,"orcid":9},"Menglong Zhao",{"name":310,"orcid":311},"Junaid Nawaz Chauhdary","https:\u002F\u002Forcid.org\u002F0000-0001-7398-5646",{"name":313,"orcid":9},"Zhaoxia Yan",{"tldr":315,"method":316,"finding":317,"direction":82,"opportunity":318},"融合PROSAIL-PRO物理模型与深度迁移学习，用无人机高光谱估算作物冠层氮含量。","PROSAIL-PRO模拟光谱预训练CNN\u002FResNet18，冻结浅层微调深层，","迁移学习模型精度与泛化性最优，小麦R²=0.86、玉米R²=0.69，缓解小样本过拟合。","可探索多作物多生育期迁移、物理模型参数不确定性传播及跨传感器泛化能力。","2026-09-15T23:30:01.518007Z"]