[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"item-3503":3,"related-3503":84},{"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":24,"tags":26,"primary_topic":32,"secondary_topics":38,"classification":39,"search_phrases":58,"slug":61,"view_count":21,"doi":62,"paper":63,"created_at":83},3503,"Adversarial Patch and Camouflage Attacks on Aerial Object Detection: A Geometry-Aware Review","https:\u002F\u002Fdoi.org\u002F10.3390\u002Fs26196057","Uncrewed aerial vehicles increasingly rely on on-board deep object detectors for applications such as surveillance, delivery, agriculture, and traffic monitoring. The vulnerability of these detectors to adversarial patches has therefore become a practical safety concern. Research on adversarial patch and camouflage attacks against aerial object detection has expanded rapidly, yet existing reviews consider aerial systems only as one application domain among many and do not examine the unique geometric conditions of aerial imagery. This review surveys adversarial patch and camouflage attacks, together with the corresponding defense mechanisms, for aerial, UAV, and remote sensing object detection. The literature is organized using a taxonomy based on two dimensions: the physical medium of the perturbation and the application domain it targets. Existing methods are then compared with respect to physical robustness, evaluation datasets and detectors, threat models, and their treatment of viewpoint variation. A recurring observation across the literature is that viewpoint variation is represented primarily through object scale, while explicit modeling of viewing geometry remains limited. This trend is closely linked to the characteristics of the aerial datasets used for evaluation, which typically provide realistic imagery but limited geometric information. Consequently, questions concerning viewpoint-dependent placement, degradation, and naturalness remain only partially explored. The review concludes by identifying the major research gaps, future directions, and deployment challenges that are likely to shape the next generation of aerial adversarial attack and defense research.","无人驾驶飞行器（UAV）在监视、配送、农业和交通监测等应用中日益依赖机载深度目标检测器。因此，这些检测器对对抗性补丁的脆弱性已成为一个实际的安全问题。针对空中目标检测的对抗性补丁与伪装攻击研究迅速扩展，然而现有综述仅将空中系统视为众多应用领域之一，并未考察航空影像独特的几何条件。本综述考察了针对空中、无人机及遥感目标检测的对抗性补丁与伪装攻击以及相应的防御机制。文献按照基于两个维度的分类体系进行组织：扰动的物理介质及其所针对的应用领域。随后，现有方法在物理鲁棒性、评估数据集与检测器、威胁模型以及其对视角变化的处理方面进行了比较。文献中反复出现的一个观察是，视角变化主要通过目标尺度来表示，而对观察几何的显式建模仍然有限。这一趋势与用于评估的航空数据集的特征密切相关，这些数据集通常提供真实影像但几何信息有限。因此，关于视角依赖的放置、退化及自然性的问题仅得到部分探讨。本综述最后指出了可能塑造下一代空中对抗攻击与防御研究的主要研究空白、未来方向及部署挑战。",null,"Sensors","2026-09-24T00:00:00Z","论文",10,false,67,{"impact":17,"substance":18,"depth":19,"authority":20,"freshness":21,"relevant":22,"comment":23},8,20,17,13,9,1,"系统梳理航空与无人机目标检测的对抗补丁与伪装攻击及防御，指出视角几何建模不足的研究空白，对农业遥感与无人机应用安全有参考价值。",[25],{"name":10,"url":6},[27,28,29,30,31],"无人机","农业人工智能","目标检测","遥感","对抗攻击",{"code":33,"slug":34,"name":35,"path":36,"parent_code":37},"I3","remote-sensing","遥感与农情监测","\u002Ftopic\u002Fdigital-agriculture\u002Fremote-sensing","I",[],{"status":40,"locked":14,"version":41,"method":42,"evidence":43},"classified","2026-10-22.2","rules",{"candidates":44,"agricultural_evidence":47},[45,52],{"code":33,"score":46,"supported":47,"title":48,"abstract":49,"tags":51},7,true,[],[30,50],"remote sensing",[30],{"code":53,"score":54,"supported":47,"title":55,"abstract":56,"tags":57},"I2",5,[],[29],[28,29],[59,60],"无人机 对抗补丁 目标检测","航空图像 对抗攻击","无人机对抗补丁目标检测-3503","10.3390\u002Fs26196057",{"doi":62,"openalex_id":64,"authors":65,"venue":10,"cited_by_count":75,"oa_url":6,"card":76,"direction":80,"ingested_from":82},"W7214217753",[66,69,72],{"name":67,"orcid":68},"Sandesh Shrestha","https:\u002F\u002Forcid.org\u002F0000-0001-5298-4468",{"name":70,"orcid":71},"K. T. Y. Mahima","https:\u002F\u002Forcid.org\u002F0000-0003-4975-9408",{"name":73,"orcid":74},"Asanka G. Perera","https:\u002F\u002Forcid.org\u002F0000-0003-4021-3943",0,{"tldr":77,"method":78,"finding":79,"direction":80,"opportunity":81},"综述无人机与遥感目标检测中的对抗补丁与伪装攻击及防御，并提出几何感知分类体系。","基于扰动介质与应用领域的分类法，比较物理鲁棒性、数据集、威胁模型与视角处理。","现有研究多以目标尺度代替视角变化，缺乏显式视角几何建模，因数据集几何信息有限。","农业遥感与作物表型","可构建含视角几何标注的农业航拍数据集，研究视角依赖的对抗补丁放置与自然性防御。","openalex","2026-09-25T23:30:31.219576Z",{"total":85,"page":22,"page_size":85,"items":86},6,[87,175,258,343,411,461],{"id":88,"title":89,"url":90,"summary":91,"summary_zh":92,"content":9,"source_name":93,"source_url":90,"published_at":94,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":95,"score_detail":96,"sources":101,"tags":103,"primary_topic":106,"secondary_topics":107,"classification":112,"search_phrases":148,"slug":151,"view_count":75,"doi":152,"paper":153,"created_at":174},4505,"Application of Deep Learning Techniques for Vineyard Mapping and Characterization Using Multispectral UAV Imagery: A Case Study from Northern Greece","https:\u002F\u002Fdoi.org\u002F10.5281\u002Fzenodo.23232412","Developments in unmanned aerial vehicle (UAV) technology and the rise of Deep Learning (DL) have drastically boosted the capacities of precision viticulture to monitor the environments of farms in a scalable and detailed manner. Vineyard mappings and extraction of vine-specific features are crucial for yield estimation, plant health assessment, and management decisions. This study examines the application of Deep Learning-based image analysis techniques for vineyard mapping and characterization using multispectral UAV image data (Deep Learning). The proposed approach integrates spatial and spectral information to delineate vineyard structures and capture key canopy characteristics relevant to precision agriculture. Multimodal datasets, including RGB and multispectral bands, facilitate improved discrimination between vines and surrounding elements. Geospatial analysis and data processing are further performed through ArcGIS Pro, enabling spatial validation of results, identifying extracted features, and creating appropriate interpretative thematic layers. The combination of UAV imagery, DL tools and GIS based analysis offers a strong and flexible workflow that can be adjusted for different grape cultivars, growth conditions and circumstances. Overall, the study demonstrates the potential benefits of the rigid union of UAV-based remote sensing with Deep Learning and GIS techniques to create a flexible, operational, and data-driven framework for vineyard management. Based on this, sustainable and fruitful management practices can be elaborated, thus optimizing resources and vineyard management decisions in precision viticulture,underlining the importance of such a technological integration for monitoring and managing modern vineyards.","无人机（UAV）技术的发展与深度学习（DL）的兴起，极大地提升了精准葡萄栽培以可扩展且精细的方式监测农场环境的能力。葡萄园制图与葡萄植株特异性特征的提取对于产量估算、植株健康评估和管理决策至关重要。本研究探讨了基于深度学习的图像分析技术在多光谱无人机影像数据（深度学习）用于葡萄园制图与表征中的应用。所提出的方法整合了空间与光谱信息，以 delineate 葡萄园结构并捕捉与精准农业相关的关键冠层特征。多模态数据集（包括RGB和多光谱波段）有助于更好地区分葡萄藤与周围元素。通过ArcGIS Pro进一步进行地理空间分析与数据处理，从而实现结果的空间验证、提取特征的识别以及适当的解释性专题图层的创建。无人机影像、深度学习工具与基于GIS的分析相结合，提供了一套强大且灵活的工作流程，可针对不同葡萄品种、生长条件和环境进行调整。总体而言，本研究展示了将基于无人机的遥感与深度学习和GIS技术 rigidly 结合，以创建灵活、可操作且数据驱动的葡萄园管理框架的潜在优势。在此基础上，可制定可持续且富有成效的管理实践，从而优化精准葡萄栽培中的资源与葡萄园管理决策，凸显了这种技术整合对现代葡萄园监测与管理的重要性。","Zenodo (CERN European Organization for Nuclear Research)","2026-10-08T00:00:00Z",68,{"impact":97,"substance":98,"depth":99,"authority":20,"freshness":21,"relevant":22,"comment":100},12,18,16,"希腊北部葡萄园的多光谱无人机深度学习制图案例，方法新颖但属区域性研究，对国内智慧农业有参考价值。",[102],{"name":93,"url":90},[104,27,28,30,105],"智慧农业","葡萄园管理",{"code":33,"slug":34,"name":35,"path":36,"parent_code":37},[108],{"code":53,"slug":109,"name":110,"path":111,"parent_code":37},"agricultural-ai","农业AI与大模型","\u002Ftopic\u002Fdigital-agriculture\u002Fagricultural-ai",{"status":40,"locked":14,"version":41,"method":42,"evidence":113},{"candidates":114,"agricultural_evidence":47},[115,120,127,134,141],{"code":33,"score":98,"supported":47,"title":116,"abstract":118,"tags":119},[117],"multispectral",[30,50,117],[30],{"code":53,"score":121,"supported":47,"title":122,"abstract":124,"tags":126},15,[123],"deep learning",[125,123],"深度学习",[28],{"code":128,"score":46,"supported":47,"title":129,"abstract":130,"tags":133},"I1",[],[131,132],"精准农业","precision agriculture",[104],{"code":135,"score":85,"supported":47,"title":136,"abstract":137,"tags":140},"B2",[],[138,139],"葡萄","grape",[],{"code":142,"score":85,"supported":47,"title":143,"abstract":144,"tags":147},"I4",[],[145,146],"空间分析","GIS",[],[149,150],"希腊 葡萄园 无人机 深度学习","多光谱 UAV 葡萄园 制图","希腊葡萄园无人机深度学习-4505","10.5281\u002Fzenodo.23232412",{"doi":152,"openalex_id":154,"authors":155,"venue":93,"cited_by_count":75,"oa_url":90,"card":169,"direction":80,"ingested_from":82},"W7220993218",[156,158,161,163,166],{"name":157,"orcid":9},"Dimitra Passa",{"name":159,"orcid":160},"Spyridon E. Detsikas","https:\u002F\u002Forcid.org\u002F0000-0002-3960-9361",{"name":162,"orcid":9},"Georgios Nektarios Tselos",{"name":164,"orcid":165},"George P. Petropoulos","https:\u002F\u002Forcid.org\u002F0000-0003-1442-1423",{"name":167,"orcid":168},"Antonis Litke","https:\u002F\u002Forcid.org\u002F0000-0002-7658-2559",{"tldr":170,"method":171,"finding":172,"direction":80,"opportunity":173},"用多光谱无人机影像和深度学习实现葡萄园制图与冠层特征提取。","多光谱无人机影像、RGB与多光谱融合、深度学习、ArcGIS Pro空间分析。","UAV+DL+GIS可灵活构建葡萄园精准管理的数据驱动框架。","可探索跨品种、跨生长阶段的迁移学习与轻量化模型，提升葡萄园实时监测泛化能力。","2026-10-09T23:30:25.912664Z",{"id":176,"title":177,"url":178,"summary":179,"summary_zh":180,"content":9,"source_name":181,"source_url":178,"published_at":182,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":183,"score_detail":184,"sources":188,"tags":190,"primary_topic":192,"secondary_topics":193,"classification":194,"search_phrases":224,"slug":227,"view_count":17,"doi":228,"paper":229,"created_at":257},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":98,"substance":185,"depth":98,"authority":186,"freshness":21,"relevant":22,"comment":187},22,14,"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。",[189],{"name":181,"url":178},[104,27,28,30,191],"作物氮素监测",{"code":33,"slug":34,"name":35,"path":36,"parent_code":37},[],{"status":40,"locked":14,"version":41,"method":42,"evidence":195},{"candidates":196,"agricultural_evidence":47},[197,202,206,213,220],{"code":33,"score":98,"supported":47,"title":198,"abstract":200,"tags":201},[199],"hyperspectral",[30,50,199],[30],{"code":53,"score":46,"supported":47,"title":203,"abstract":204,"tags":205},[],[125,123],[28],{"code":207,"score":85,"supported":47,"title":208,"abstract":209,"tags":212},"A2",[],[210,211],"小麦","wheat",[],{"code":214,"score":85,"supported":47,"title":215,"abstract":216,"tags":219},"A3",[],[217,218],"玉米","maize",[],{"code":128,"score":22,"supported":14,"title":221,"abstract":222,"tags":223},[],[],[104],[225,226],"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":228,"openalex_id":230,"authors":231,"venue":181,"cited_by_count":75,"oa_url":178,"card":252,"direction":80,"ingested_from":82},"W7212964562",[232,234,236,238,240,243,245,247,250],{"name":233,"orcid":9},"Jing Zhao",{"name":235,"orcid":9},"Hong Li",{"name":237,"orcid":9},"Junping Liu",{"name":239,"orcid":9},"Wei Chen",{"name":241,"orcid":242},"Xin Guo","https:\u002F\u002Forcid.org\u002F0000-0002-1583-5364",{"name":244,"orcid":9},"Yunlong Wu",{"name":246,"orcid":9},"Menglong Zhao",{"name":248,"orcid":249},"Junaid Nawaz Chauhdary","https:\u002F\u002Forcid.org\u002F0000-0001-7398-5646",{"name":251,"orcid":9},"Zhaoxia Yan",{"tldr":253,"method":254,"finding":255,"direction":80,"opportunity":256},"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",{"id":259,"title":260,"url":261,"summary":262,"summary_zh":263,"content":9,"source_name":264,"source_url":261,"published_at":265,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":266,"score_detail":267,"sources":269,"tags":271,"primary_topic":273,"secondary_topics":279,"classification":282,"search_phrases":304,"slug":307,"view_count":308,"doi":309,"paper":310,"created_at":342},2320,"Genotype-Aware Prediction of Soybean Seed Composition from Multimodal UAV Imagery of the Standing Crop","https:\u002F\u002Fdoi.org\u002F10.3390\u002Frs18183121","Geospatial artificial intelligence (GeoAI) integrates multimodal remote sensing with deep learning to model complex agricultural systems at scale. Within this framework, accurate and non-destructive prediction of seed composition from in-season standing crops is essential for breeding and precision agriculture. This study developed an end-to-end convolutional neural network (CNN) framework to estimate eight seed traits (protein, oil, sucrose, fiber, starch, ash, complex and simple carbohydrates) from UAV-based multisensor imagery and associated genotype and phenological metadata. A total of 372 soybean samples were collected over two growing seasons (2020–2021) from two fields in Missouri, with UAV flights capturing multispectral (MSI), thermal (THR), and LiDAR (LDR) data at four time points spanning vegetative to reproductive growth stages. CNN models were trained in single- and multi-date configurations, incorporating genotype (GEN) and days after sowing (DAS) as additional features. The highest accuracy was achieved for sucrose (R2 = 0.80), followed by simple carbohydrate (R2 = 0.70) and starch (R2 = 0.55), with notable gains from GEN and DAS. Multi-date models incorporating earlier acquisitions often matched or outperformed later or all-date combinations. Among modalities, MSI provided the most robust estimates, with limited added value from LDR or THR. Unlike feature-based pipelines prone to multicollinearity, this image-to-trait approach enables automated, scalable prediction of soybean seed composition for in-season, field-level assessment.","地理空间人工智能（GeoAI）将多模态遥感与深度学习相结合，以在大尺度上对复杂农业系统进行建模。在该框架下，对当季未收获作物进行准确且无损的种子成分预测，对于育种和精准农业至关重要。本研究开发了一种端到端的卷积神经网络（CNN）框架，利用基于无人机（UAV）的多传感器影像以及相关的基因型和物候元数据，估算八种种子性状（蛋白质、油分、蔗糖、纤维、淀粉、灰分、复杂碳水化合物和简单碳水化合物）。研究在2020—2021两个生长季内，从密苏里州两块田地共采集372份大豆样本，无人机飞行在从营养生长期到生殖生长期的四个时间点获取了多光谱（MSI）、热红外（THR）和激光雷达（LDR）数据。CNN模型以单日期和多日期配置进行训练，并将基因型（GEN）和播种后天数（DAS）作为附加特征纳入。蔗糖的估算精度最高（R² = 0.80），其次为简单碳水化合物（R² = 0.70）和淀粉（R² = 0.55），且GEN和DAS的加入带来了显著提升。纳入早期采集数据的多日期模型往往与后期或全日期组合的表现相当，甚至更优。在各模态中，MSI提供了最稳健的估算结果，而LDR或THR的附加价值有限。与易受多重共线性影响的特征工程流程不同，这种从图像到性状的方法能够实现大豆种子成分的自动化、可扩展预测，用于当季田块尺度的评估。","Remote Sensing","2026-09-11T00:00:00Z",80,{"impact":98,"substance":185,"depth":98,"authority":186,"freshness":17,"relevant":22,"comment":268},"基于多模态无人机影像与基因型信息的大豆籽粒成分无损预测研究，方法新颖、数据扎实，对智慧育种与精准农业有参考价值。",[270],{"name":264,"url":261},[104,27,28,30,272],"大豆育种",{"code":274,"slug":275,"name":276,"path":277,"parent_code":278},"A4","soybean-oilseeds","大豆与油料","\u002Ftopic\u002Ffield-crops\u002Fsoybean-oilseeds","A",[280,281],{"code":53,"slug":109,"name":110,"path":111,"parent_code":37},{"code":33,"slug":34,"name":35,"path":36,"parent_code":37},{"status":40,"locked":14,"version":41,"method":42,"evidence":283},{"candidates":284,"agricultural_evidence":47},[285,291,296,300],{"code":274,"score":121,"supported":47,"title":286,"abstract":288,"tags":290},[287],"soybean",[289,287],"大豆",[272],{"code":53,"score":13,"supported":47,"title":292,"abstract":293,"tags":295},[],[125,123,294],"artificial intelligence",[28],{"code":33,"score":13,"supported":47,"title":297,"abstract":298,"tags":299},[],[30,50,117],[30],{"code":128,"score":46,"supported":47,"title":301,"abstract":302,"tags":303},[],[131,132],[104],[305,306],"农业人工智能 大豆育种 智慧农业 无人机","农业人工智能 大豆育种","农业人工智能大豆育种智慧农业无人机-2320",4,"10.3390\u002Frs18183121",{"doi":309,"openalex_id":311,"authors":312,"venue":264,"cited_by_count":75,"oa_url":261,"card":337,"direction":80,"ingested_from":82},"W7212223520",[313,316,319,322,325,327,329,332,335],{"name":314,"orcid":315},"Vasit Sagan","https:\u002F\u002Forcid.org\u002F0000-0003-4375-2096",{"name":317,"orcid":318},"Kristen Rhodes","https:\u002F\u002Forcid.org\u002F0000-0002-8469-5043",{"name":320,"orcid":321},"Sourav Bhadra","https:\u002F\u002Forcid.org\u002F0000-0002-5832-4695",{"name":323,"orcid":324},"Haireti Alifu","https:\u002F\u002Forcid.org\u002F0000-0002-7369-6657",{"name":326,"orcid":9},"Aviskar Giri",{"name":328,"orcid":9},"Ashutosh Pawar",{"name":330,"orcid":331},"Bishal Roy","https:\u002F\u002Forcid.org\u002F0000-0001-9912-5505",{"name":333,"orcid":334},"Supria Sarkar","https:\u002F\u002Forcid.org\u002F0000-0001-9467-5786",{"name":336,"orcid":9},"Felix Fritschi",{"tldr":338,"method":339,"finding":340,"direction":80,"opportunity":341},"用无人机多模态影像与基因型数据训练CNN，预测大豆八种种子成分。","372份大豆样本，多光谱\u002F热红外\u002FLiDAR影像加基因型与播后天数，端到端CNN","蔗糖预测最优(R²=0.80)，基因型与播期提升精度，多光谱最稳健。","可探索基因型-表型-环境交互建模，并迁移至其他作物与多时相早期预测。","2026-09-13T23:30:22.776177Z",{"id":344,"title":345,"url":346,"summary":347,"summary_zh":348,"content":9,"source_name":264,"source_url":346,"published_at":349,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":350,"score_detail":351,"sources":353,"tags":355,"primary_topic":358,"secondary_topics":359,"classification":361,"search_phrases":377,"slug":380,"view_count":381,"doi":382,"paper":383,"created_at":410},1734,"Lightweight Near-Infrared Spectral Reconstruction from Red UAV Imagery Using Artificial Intelligence for Low-Cost Remote Sensing","https:\u002F\u002Fdoi.org\u002F10.3390\u002Frs18173015","Near-infrared imagery is essential for vegetation monitoring, precision agriculture, and environmental remote sensing, but multispectral UAV systems remain significantly more expensive and less accessible than conventional RGB imaging platforms. This study presents a lightweight artificial intelligence framework for reconstructing the NIR spectral band exclusively from the red spectral band acquired by a UAV. The proposed methodology formulates the reconstruction task as a pixel-wise nonlinear regression problem and employs a compact multilayer perceptron (MLP) containing only 609 trainable parameters, without exploiting spatial neighborhood information. The framework was developed and evaluated using 280 synchronized multispectral UAV image sets acquired with a DJI Phantom 4 Multispectral platform over a heterogeneous agricultural landscape in the Republic of Moldova. Of these, 252 image sets were used for model development, and 28 were reserved as a held-out within-mission test subset. Quantitative evaluation on a held-out test dataset from the same acquisition mission yielded a mean squared error of 0.010329, a root mean squared error of 0.101632, a mean absolute error of 0.079883, a coefficient of determination of 0.253383, and a Pearson correlation coefficient of 0.683637 between measured and reconstructed normalized NIR digital intensities. The results indicate that the model captures part of the red–NIR relationship under the evaluated acquisition conditions; however, the moderate coefficient of determination suggests that the reconstructed values are an approximation rather than a replacement for measured NIR observations. An illustrative NDVI-based assessment showed that broad spatial vegetation patterns remained identifiable. Rather than introducing a new neural network architecture, this work establishes a compact empirical baseline to investigate the practical performance and limitations of pixel-wise NIR reconstruction from a single red-band value with minimal model complexity.","近红外影像对于植被监测、精准农业和环境遥感至关重要，但多光谱无人机系统仍比传统RGB成像平台昂贵得多，且可及性较低。本研究提出了一种轻量级人工智能框架，仅从无人机获取的红色波段重建近红外光谱波段。所提方法将重建任务表述为逐像素非线性回归问题，并采用仅含609个可训练参数的紧凑型多层感知器（MLP），不利用空间邻域信息。该框架基于在摩尔多瓦共和国异质农业景观上使用DJI Phantom 4 Multispectral平台采集的280组同步多光谱无人机影像集进行开发和评估。其中，252组影像集用于模型开发，28组作为任务内保留测试子集。对来自同一采集任务的保留测试数据集进行的定量评估显示，实测与重建的归一化近红外数字强度之间的均方误差为0.010329，均方根误差为0.101632，平均绝对误差为0.079883，决定系数为0.253383，皮尔逊相关系数为0.683637。结果表明，在评估的采集条件下，模型捕捉了红–近红外关系的一部分；然而，中等的决定系数表明，重建值是对实测近红外观测的近似，而非替代。一项基于NDVI的示例评估显示，大尺度空间植被模式仍可辨识。本研究并非引入新的神经网络架构，而是建立了一个紧凑的经验基线，以在最小模型复杂度下探究基于单一红色波段值进行逐像素近红外重建的实际性能与局限性。","2026-09-04T00:00:00Z",58,{"impact":17,"substance":98,"depth":99,"authority":97,"freshness":308,"relevant":22,"comment":352},"研究用AI从红波段重建近红外，降低遥感成本，方法新颖但精度有限，对低成本农业监测有参考价值。",[354],{"name":264,"url":346},[27,28,30,356,357],"植被监测","近红外",{"code":33,"slug":34,"name":35,"path":36,"parent_code":37},[360],{"code":53,"slug":109,"name":110,"path":111,"parent_code":37},{"status":40,"locked":14,"version":41,"method":42,"evidence":362},{"candidates":363,"agricultural_evidence":47},[364,369,373],{"code":33,"score":365,"supported":47,"title":366,"abstract":367,"tags":368},19,[50],[30,50,117],[356,30],{"code":53,"score":97,"supported":47,"title":370,"abstract":371,"tags":372},[294],[294],[28],{"code":128,"score":85,"supported":47,"title":374,"abstract":375,"tags":376},[],[131,132],[],[378,379],"农业人工智能 植被监测 无人机 近红外","农业人工智能 植被监测","农业人工智能植被监测无人机近红外-1734",3,"10.3390\u002Frs18173015",{"doi":382,"openalex_id":384,"authors":385,"venue":264,"cited_by_count":75,"oa_url":346,"card":405,"direction":80,"ingested_from":82},"W7208716610",[386,389,391,394,396,399,402],{"name":387,"orcid":388},"Viorel Bostan","https:\u002F\u002Forcid.org\u002F0000-0002-2422-3538",{"name":390,"orcid":9},"Nicu Drumea",{"name":392,"orcid":393},"Viorel Cărbune","https:\u002F\u002Forcid.org\u002F0000-0002-1556-4453",{"name":395,"orcid":9},"Valeriu Seinic",{"name":397,"orcid":398},"Igor Calmîcov","https:\u002F\u002Forcid.org\u002F0000-0002-4017-3702",{"name":400,"orcid":401},"Adriana Ursu","https:\u002F\u002Forcid.org\u002F0009-0003-4263-2444",{"name":403,"orcid":404},"Maria Gutu","https:\u002F\u002Forcid.org\u002F0000-0002-2820-393X",{"tldr":406,"method":407,"finding":408,"direction":80,"opportunity":409},"用仅含609个参数的MLP从UAV红光波段重建近红外波段，低成本替代多光谱相机。","像素级非线性回归，紧凑MLP（609参数），280组多光谱UAV图像训练。","重建NIR与实测相关性中等（R²=0.253），可识别植被模式，但精度有限。","可探索引入空间信息或时序特征提升重建精度，或扩展至其他波段与作物类型。","2026-09-05T23:30:27.438287Z",{"id":412,"title":413,"url":414,"summary":415,"summary_zh":416,"content":9,"source_name":93,"source_url":414,"published_at":417,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":15,"score_detail":418,"sources":420,"tags":422,"primary_topic":425,"secondary_topics":426,"classification":428,"search_phrases":445,"slug":448,"view_count":75,"doi":449,"paper":450,"created_at":460},5570,"Evaluation of Counterfactual Explanation Robustness for Computer Vision Agriculture Yield-Forecast Models Under Satellite Cloud-Mask Gaps: An Interpretability Reliability Bound for Cooperative Storage Decisions","https:\u002F\u002Fdoi.org\u002F10.5281\u002Fzenodo.23545079","Counterfactual explanations are increasingly proposed as the interface through which cooperative storage managers interrogate computer-vision yield-forecast models, yet their reliability under the satellite cloud-mask gaps that routinely afflict agricultural remote sensing has not been bounded. The bound is derived from a Lipschitz continuity assumption on the forecast model and a masking-operator perturbation analysis, and it is expressed as a function of the cloud-mask gap fraction, the model's local Lipschitz constant, and the counterfactual search tolerance. A replication-based evaluation of the before-after-control-impact estimator was conducted to characterise the operating characteristics of the design under a declared variance structure (between-cluster SD = 6.0, residual SD = 4.5, intraclass correlation = 0.64), with power estimated across 2000 design cells. The reliability bound degrades non-linearly as the gap fraction rises, and that the intraclass correlation of 0.64 implies that cooperative-level aggregation of counterfactual recourse is necessary to preserve explanation stability. The contribution is a decision-relevant reliability bound that storage cooperatives can use to set a maximum tolerable cloud-mask gap before counterfactual recourse is suspended.","反事实解释日益被提出作为合作式仓储管理者质询计算机视觉产量预测模型的接口，然而在农业遥感中经常出现的卫星云掩膜缺口条件下，其可靠性尚未得到界定。该界基于预测模型上的Lipschitz连续性假设和掩膜算子扰动分析推导得出，并表示为云掩膜缺口比例、模型局部Lipschitz常数和反事实搜索容差的函数。对前后对照影响估计量进行了基于重复的评估，以表征该设计在给定方差结构（组间标准差=6.0，残差标准差=4.5，组内相关系数=0.64）下的运行特征，并在2000个设计单元中估计了检验效能。可靠性界随缺口比例上升呈非线性退化，且组内相关系数0.64意味着需要对反事实救济进行合作社层面的聚合，以保持解释的稳定性。本文的贡献在于提供了一个与决策相关的可靠性界，仓储合作社可据此设定在暂停反事实救济之前可容忍的最大云掩膜缺口。","2026-10-21T00:00:00Z",{"impact":17,"substance":18,"depth":19,"authority":97,"freshness":13,"relevant":22,"comment":419},"该论文提出反事实解释在卫星云掩膜缺口下的可靠性边界，方法新颖且结论可用于合作社决策，但属细分方法学研究，产业影响有限。",[421],{"name":93,"url":414},[104,28,423,30,424],"产量预测","可解释性",{"code":53,"slug":109,"name":110,"path":111,"parent_code":37},[427],{"code":33,"slug":34,"name":35,"path":36,"parent_code":37},{"status":40,"locked":14,"version":41,"method":42,"evidence":429},{"candidates":430,"agricultural_evidence":47},[431,437,441],{"code":53,"score":97,"supported":47,"title":432,"abstract":434,"tags":436},[433],"computer vision",[435],"计算机视觉",[28],{"code":33,"score":17,"supported":47,"title":438,"abstract":439,"tags":440},[],[30,50],[423,30],{"code":128,"score":22,"supported":14,"title":442,"abstract":443,"tags":444},[],[],[104],[446,447],"农业人工智能 产量预测 可解释性 智慧农业","农业人工智能 产量预测","农业人工智能产量预测可解释性智慧农业-5570","10.5281\u002Fzenodo.23545079",{"doi":449,"openalex_id":451,"authors":452,"venue":93,"cited_by_count":75,"oa_url":414,"card":455,"direction":80,"ingested_from":82},"W7223114002",[453],{"name":454,"orcid":9},"Kadiatou Barry",{"tldr":456,"method":457,"finding":458,"direction":80,"opportunity":459},"推导卫星云掩膜缺口下反事实解释可靠性的理论界，并评估其对合作社储粮决策的影响。","基于Lipschitz连续性假设与掩膜扰动分析推导可靠性界，并用2000个设计单","缺口比例上升使可靠性非线性下降；ICC=0.64表明需合作社层级聚合以保持解释稳定。","可研究不同作物区云掩膜统计特性下的自适应解释暂停阈值与聚合策略。","2026-10-22T11:20:39.240791Z",{"id":462,"title":463,"url":464,"summary":465,"summary_zh":466,"content":9,"source_name":467,"source_url":464,"published_at":468,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":469,"score_detail":470,"sources":472,"tags":474,"primary_topic":477,"secondary_topics":482,"classification":484,"search_phrases":506,"slug":509,"view_count":75,"doi":510,"paper":511,"created_at":553},5567,"Capturing Spatial Heterogeneity Using UAV-Derived Multispectral Phenotypes and Random Regression Models Improves Genetic Evaluation in Alfalfa Breeding Trials","https:\u002F\u002Fdoi.org\u002F10.64898\u002F2026.10.14.759249","Spatial heterogeneity is a major source of environmental noise in field trials, reducing genetic evaluation accuracy. Conventional spatial correction methods use field layout and plot proximity but often fail to capture in-season dynamics or leverage temporal data from high-throughput phenotyping. We hypothesized that longitudinal UAV-derived multispectral measurements could characterize field-scale environmental heterogeneity and improve genetic evaluation in breeding trials. Multispectral imagery was collected throughout the growing season across eight alfalfa trials in New York and New Mexico (20222023). Vegetation indices were evaluated for heritability and association with biomass yield using univariate and multi-trait models. Longitudinal measurements were modeled with random regression (RR) to estimate plot-specific permanent environmental effects (PEE), incorporated into genomic mixed models within a two-stage framework. Model performance was compared with conventional autoregressive (AR) and baseline (no correction) models using metrics of consistency and accuracy of estimated genetic effects across replicates. Vegetation indices showed moderate to high heritability, with GNDVI consistently outperforming NDVI and NDRE. Both univariate and multi-trait analyses revealed substantial spatial variability, particularly in New Mexico trials. RR-derived PEE captured field-scale spatial patterns and improved genetic signal recovery relative to baseline and AR models, with the largest gains under strong spatial heterogeneity and comparable performance where variability was limited. The RR framework improved reproducibility of genetic effect estimates across independent field subsets without compromising agreement with full-trial evaluations. These findings show that integrating UAV data via random regression enhances genetic signal recovery and selection accuracy, with broader potential for precision agriculture and biosensor-driven crop management.","空间异质性是田间试验中环境噪声的主要来源，会降低遗传评估的准确性。传统空间校正方法利用田间布局和小区邻近性，但往往无法捕捉生育期内的动态变化，也未能充分利用高通量表型分析产生的时间序列数据。我们假设，基于无人机（UAV）的纵向多光谱测量能够表征田块尺度的环境异质性，并改善育种试验中的遗传评估。在2022—2023年纽约州和新墨西哥州的8个紫花苜蓿试验中，于整个生长季收集了多光谱影像。采用单性状和多性状模型评估了植被指数的遗传力及其与生物量产量的关联。纵向测量数据采用随机回归（RR）建模，以估计小区特异性永久环境效应（PEE），并在两阶段框架中将其纳入基因组混合模型。利用重复间估计遗传效应的一致性和准确性指标，将模型性能与传统的自回归（AR）模型和基线（无校正）模型进行比较。植被指数表现出中到高的遗传力，其中GNDVI始终优于NDVI和NDRE。单性状和多性状分析均揭示了显著的空间变异，尤其在新墨西哥州的试验中更为明显。RR衍生的PEE捕捉到了田块尺度的空间格局，相对于基线模型和AR模型改善了遗传信号的恢复，在强空间异质性条件下增益最大，而在变异有限时性能相当。RR框架提高了遗传效应估计在独立田间子集间的可重复性，同时未损害与全试验评估的一致性。这些结果表明，通过随机回归整合UAV数据可增强遗传信号恢复和选择准确性，在精准农业和生物传感器驱动的作物管理方面具有更广泛的潜力。","bioRxiv (Cold Spring Harbor Laboratory)","2026-10-20T00:00:00Z",79,{"impact":98,"substance":185,"depth":98,"authority":97,"freshness":21,"relevant":22,"comment":471},"该研究利用无人机多光谱表型与随机回归模型校正田间空间异质性，显著提升苜蓿育种遗传评估精度，方法新颖、数据覆盖两州八组试验，对智慧育种有实质参考价值。",[473],{"name":467,"url":464},[104,27,475,30,476],"育种","紫花苜蓿",{"code":478,"slug":479,"name":480,"path":481,"parent_code":278},"A7","forage-cropping","饲草与耕作制度","\u002Ftopic\u002Ffield-crops\u002Fforage-cropping",[483],{"code":33,"slug":34,"name":35,"path":36,"parent_code":37},{"status":40,"locked":14,"version":41,"method":42,"evidence":485},{"candidates":486,"agricultural_evidence":47},[487,493,497,501],{"code":478,"score":121,"supported":47,"title":488,"abstract":490,"tags":492},[489],"alfalfa",[491,489],"苜蓿",[476],{"code":33,"score":97,"supported":47,"title":494,"abstract":495,"tags":496},[117],[117],[30],{"code":128,"score":46,"supported":47,"title":498,"abstract":499,"tags":500},[],[131,132],[104],{"code":502,"score":308,"supported":47,"title":503,"abstract":504,"tags":505},"C4",[],[491],[476],[507,508],"无人机 多光谱 苜蓿 育种","随机回归 空间异质性 遗传评估","无人机多光谱苜蓿育种-5567","10.64898\u002F2026.10.14.759249",{"doi":510,"openalex_id":512,"authors":513,"venue":467,"cited_by_count":75,"oa_url":464,"card":548,"direction":80,"ingested_from":82},"W7223083966",[514,517,520,523,526,528,530,533,536,539,542,545],{"name":515,"orcid":516},"Dinesh Ghimire","https:\u002F\u002Forcid.org\u002F0009-0008-2281-1396",{"name":518,"orcid":519},"Ranjita Thapa","https:\u002F\u002Forcid.org\u002F0000-0001-6116-7611",{"name":521,"orcid":522},"Karl H. Kunze","https:\u002F\u002Forcid.org\u002F0000-0003-4548-1808",{"name":524,"orcid":525},"Felipe Sabadin","https:\u002F\u002Forcid.org\u002F0000-0003-2937-1465",{"name":527,"orcid":9},"Julie Hansen",{"name":529,"orcid":9},"Christopher Pierce",{"name":531,"orcid":532},"Liam Wickes-Do","https:\u002F\u002Forcid.org\u002F0009-0003-2927-4590",{"name":534,"orcid":535},"Nicholas Santantonio","https:\u002F\u002Forcid.org\u002F0000-0002-4351-4023",{"name":537,"orcid":538},"Ian M. Ray","https:\u002F\u002Forcid.org\u002F0000-0003-2056-536X",{"name":540,"orcid":541},"Michael A. Gore","https:\u002F\u002Forcid.org\u002F0000-0001-6896-8024",{"name":543,"orcid":544},"Virginia Mae Moore","https:\u002F\u002Forcid.org\u002F0000-0001-7888-3366",{"name":546,"orcid":547},"Kelly R. Robbins","https:\u002F\u002Forcid.org\u002F0000-0001-9522-9585",{"tldr":549,"method":550,"finding":551,"direction":80,"opportunity":552},"用无人机多光谱时序表型和随机回归模型校正田间空间异质性，提升苜蓿育种遗传评估精度。","八项苜蓿试验的无人机多光谱植被指数，随机回归估计小区永久环境效应并嵌入基因组混合","随机回归提取的环境效应能捕捉田块尺度空间格局，在强异质性下显著提升遗传信号恢复与选择准确性。","可将无人机时序表型与随机回归框架推广到其他作物育种试验，并探索多源遥感与基因组选择的联合建模。","2026-10-22T11:20:32.655394Z"]