[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"item-3829":3,"related-3829":109},{"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,"primary_topic":31,"secondary_topics":37,"classification":44,"search_phrases":80,"slug":83,"view_count":84,"doi":85,"paper":86,"created_at":108},3829,"Towards accessible smallholder crop classification in the groundnut basin of Senegal","https:\u002F\u002Fdoi.org\u002F10.1088\u002F2976-601x\u002Faea0b8","Abstract Crop type mapping is a critical lever for decision making in smallholder food systems, yet high cost and expertise requirements of developing remote sensing-driven maps as well as need for recurrent data are persistent barriers to incorporating geospatial tools into decision making. To address this gap, we establish a three-part framework to evaluate crop type mapping: (1) performance, (2) temporal transferability, and (3) accessibility. We evaluate geospatial foundation model embeddings (TESSERA and AlphaEarth) against traditional baselines in Senegal’s groundnut basin. We find that the embedding-based approaches, in particular TESSERA, best fulfill these criteria. The TESSERA-based method shows a 78% reduction in both compute compared to a timeseries approach, a reduction in need for custom feature generation, and 28% higher relative accuracy (an absolute improvement of approximately 0.10 weighted F 1) in one crop type classification temporal transfer test. By reducing the reliance on training data, computational resource use, and expertise for data processing, this approach shows potential to contribute to improved smallholder crop type classification.","摘要 作物类型制图是小农粮食系统决策的关键手段，但开发遥感驱动地图的高成本与专业知识要求，以及 recurring 数据的需求，仍是将地理空间工具纳入决策的持续障碍。为弥补这一差距，我们建立了一个三部分框架来评估作物类型制图：（1）性能，（2）时间可迁移性，（3）可及性。我们在塞内加尔花生盆地评估了地理空间基础模型嵌入（TESSERA 和 AlphaEarth）与传统基线方法。我们发现，基于嵌入的方法，尤其是 TESSERA，最能满足这些标准。基于 TESSERA 的方法显示，与时间序列方法相比，计算量减少 78%，对自定义特征生成的需求降低，并且在一项作物类型分类时间迁移测试中相对精度提高 28%（加权 F1 绝对提升约 0.10）。通过减少对训练数据、计算资源使用和数据处理专业知识的依赖，该方法显示出改善小农作物类型分类的潜力。",null,"Environmental Research Food Systems","2026-09-28T00:00:00Z","论文",10,false,79,{"impact":17,"substance":18,"depth":17,"authority":19,"freshness":20,"relevant":21,"comment":22},18,22,13,8,1,"该研究提出兼顾性能、时序迁移性与可及性的作物分类评估框架，用地理空间基础模型显著降低算力与训练数据依赖，对小农户农业遥感应用具有实质参考价值。",[24],{"name":10,"url":6},[26,27,28,29,30],"智慧农业","农业人工智能","农业遥感","小农户","作物分类",{"code":32,"slug":33,"name":34,"path":35,"parent_code":36},"G2","farm-services","社会化服务与规模经营","\u002Ftopic\u002Frural-services\u002Ffarm-services","G",[38],{"code":39,"slug":40,"name":41,"path":42,"parent_code":43},"I3","remote-sensing","遥感与农情监测","\u002Ftopic\u002Fdigital-agriculture\u002Fremote-sensing","I",{"status":45,"locked":14,"version":46,"method":47,"evidence":48},"classified","2026-10-22.2","rules",{"candidates":49,"agricultural_evidence":52},[50,57,63,70,75],{"code":32,"score":51,"supported":52,"title":53,"abstract":55,"tags":56},12,true,[54],"smallholder",[54],[29],{"code":39,"score":20,"supported":52,"title":58,"abstract":59,"tags":62},[],[60,61],"遥感","remote sensing",[30,28],{"code":64,"score":65,"supported":52,"title":66,"abstract":67,"tags":69},"A4",3,[],[68],"花生",[],{"code":71,"score":21,"supported":14,"title":72,"abstract":73,"tags":74},"I1",[],[],[26],{"code":76,"score":21,"supported":14,"title":77,"abstract":78,"tags":79},"I2",[],[],[27],[81,82],"塞内加尔 花生盆地 作物分类","TESSERA 遥感 作物制图","塞内加尔花生盆地作物分类-3829",5,"10.1088\u002F2976-601x\u002Faea0b8",{"doi":85,"openalex_id":87,"authors":88,"venue":10,"cited_by_count":98,"oa_url":99,"card":100,"direction":106,"ingested_from":107},"W7214087594",[89,91,93,95],{"name":90,"orcid":9},"Madeline Lisaius",{"name":92,"orcid":9},"Srinivasan Keshav",{"name":94,"orcid":9},"Andrew Blake",{"name":96,"orcid":97},"Clement Atzberger","https:\u002F\u002Forcid.org\u002F0000-0003-2169-8009",0,"https:\u002F\u002Fiopscience.iop.org\u002Farticle\u002F10.1088\u002F2976-601X\u002Faea0b8\u002Fpdf",{"tldr":101,"method":102,"finding":103,"direction":104,"opportunity":105},"评估地理空间基础模型嵌入在塞内加尔花生盆地小农作物分类中的性能、时序迁移性和可及性。","对比TESSERA和AlphaEarth嵌入与传统基线，用加权F1和计算量评估。","TESSERA嵌入精度相对提高28%，计算量减少78%，且无需自定义特征生成。","农业遥感与作物表型","可探索基础模型嵌入在更多小农区域和作物类型上的泛化能力及低成本部署方案。","智慧农业 \u002F 农业物联网","openalex","2026-09-30T23:30:11.490990Z",{"total":110,"page":21,"page_size":110,"items":111},6,[112,222,282,335,428,498],{"id":113,"title":114,"url":115,"summary":116,"summary_zh":117,"content":9,"source_name":118,"source_url":115,"published_at":119,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":120,"score_detail":121,"sources":125,"tags":127,"primary_topic":129,"secondary_topics":130,"classification":131,"search_phrases":184,"slug":187,"view_count":65,"doi":188,"paper":189,"created_at":221},4091,"Supporting Digital Agriculture Through GeoAI : A Crop Type Dataset of Four Food Basket Crops in Rwanda","https:\u002F\u002Fdoi.org\u002F10.1002\u002Fgdj3.70097","ABSTRACT Agriculture is a major economic activity, contributing approximately 25% of Rwanda's gross domestic product and employing nearly 65% of its population. Recognising its critical role, the Government of Rwanda has prioritised digital agriculture to manage the sector sustainably. Decision support systems based on remote sensing and Geospatial Artificial Intelligence (GeoAI) offer a transformative opportunity to enhance regional food security by providing real‐time data, improving agricultural forecasting and mitigating the impacts of climate variability. However, GeoAI approaches require labelled, ground‐referenced crop type data for model calibration and validation, which is scarce in sub‐Saharan Africa. To address this gap, we collected crop type data across Musanze, Nyabihu, Nyagatare and Ruhango districts of Rwanda between May and June 2025 to support the development of GeoAI crop monitoring approaches. The dataset, collected using KoboToolbox, integrates the Joint Experiment for Crop Assessment and Monitoring guidelines for field data collection and the Food and Agriculture Organisation's indicative crop classification. We targeted four food‐basket crops from Rwanda's Crop Intensification Programme: bean, maize, Irish potato and rice. In total, 2739 GPS‐located points were collected and used to generate 914 crop‐labelled polygons: 112 for beans, 107 for Irish potato, 417 for maize, 34 for rice and 244 for mixed‐crop fields. Polygons were digitised using four cardinal‐direction photographs taken during fieldwork and Planet imagery acquired during the season. Besides crop labels, the point data include detailed agronomic information such as farm management practices, anti‐erosion measures, cropping systems, agroforestry practices (including fruit‐tree presence), administrative units and GPS coordinates. This rich dataset provides robust, nuanced information for crop monitoring, yield estimation and digital agriculture innovation in Rwanda and beyond.","农业是一项主要的经济活动，贡献了卢旺达约25%的国内生产总值，并雇用了近65%的人口。认识到农业的关键作用，卢旺达政府已将数字农业列为优先事项，以实现该部门的可持续管理。基于遥感和地理空间人工智能（GeoAI）的决策支持系统提供了变革性机遇，可通过提供实时数据、改进农业预测和减轻气候变率的影响来增强区域粮食安全。然而，GeoAI方法需要标记的地面参考作物类型数据用于模型校准和验证，而这在撒哈拉以南非洲地区十分稀缺。为填补这一空白，我们于2025年5月至6月期间在卢旺达穆桑泽、尼亚比胡、尼亚加塔雷和鲁汉戈四个地区收集了作物类型数据，以支持GeoAI作物监测方法的开发。该数据集使用KoboToolbox收集，整合了作物评估与监测联合实验的野外数据收集指南以及联合国粮食及农业组织的指示性作物分类。我们针对卢旺达作物集约化计划中的四种粮篮作物：菜豆、玉米、爱尔兰马铃薯和水稻。共收集了2739个GPS定位点，用于生成914个作物标记多边形：112个菜豆、107个爱尔兰马铃薯、417个玉米、34个水稻和244个混合作物田块。多边形利用野外工作中拍摄的四个基本方位照片和当季获取的Planet影像进行数字化。除作物标签外，点位数据还包括详细的农艺信息，如农场管理实践、防侵蚀措施、种植制度、农林业实践（包括果树存在情况）、行政单元和GPS坐标。这一丰富的数据集为卢旺达及其他地区的作物监测、产量估算和数字农业创新提供了稳健而细致的信息。","Geoscience Data Journal","2026-10-01T00:00:00Z",78,{"impact":122,"substance":18,"depth":17,"authority":123,"freshness":20,"relevant":21,"comment":124},16,14,"基于GeoAI的卢旺达四类主粮作物地面标注数据集，2739个GPS点位、914个作物多边形，数据规模与方法规范性强，对遥感作物监测与数字农业研究有实质参考价值。",[126],{"name":118,"url":115},[128,26,27,28,30],"数字农业",{"code":39,"slug":40,"name":41,"path":42,"parent_code":43},[],{"status":45,"locked":14,"version":46,"method":47,"evidence":132},{"candidates":133,"agricultural_evidence":52},[134,139,143,150,157,164,171,178],{"code":71,"score":122,"supported":52,"title":135,"abstract":137,"tags":138},[136],"digital agriculture",[128,136],[128,26],{"code":39,"score":20,"supported":52,"title":140,"abstract":141,"tags":142},[],[60,61],[30,28],{"code":76,"score":144,"supported":52,"title":145,"abstract":146,"tags":149},7,[],[147,148],"artificial intelligence","decision support",[27],{"code":151,"score":110,"supported":52,"title":152,"abstract":153,"tags":156},"A1",[],[154,155],"水稻","rice",[],{"code":158,"score":110,"supported":52,"title":159,"abstract":160,"tags":163},"A3",[],[161,162],"玉米","maize",[],{"code":165,"score":110,"supported":52,"title":166,"abstract":167,"tags":170},"A5",[],[168,169],"马铃薯","potato",[],{"code":172,"score":110,"supported":52,"title":173,"abstract":174,"tags":177},"A8",[],[175,176],"粮食安全","food security",[],{"code":179,"score":65,"supported":52,"title":180,"abstract":181,"tags":183},"B2",[],[182],"果树",[],[185,186],"卢旺达 作物类型 数据集","GeoAI 作物监测 遥感","卢旺达作物类型数据集-4091","10.1002\u002Fgdj3.70097",{"doi":188,"openalex_id":190,"authors":191,"venue":118,"cited_by_count":98,"oa_url":115,"card":216,"direction":104,"ingested_from":107},"W7215812020",[192,195,197,200,202,204,206,209,211,213],{"name":193,"orcid":194},"Benson Kipkemboi Kenduiywo","https:\u002F\u002Forcid.org\u002F0000-0002-8448-0499",{"name":196,"orcid":9},"Stephen Sande",{"name":198,"orcid":199},"Anastasia Mumbi Wahome","https:\u002F\u002Forcid.org\u002F0000-0002-0471-0697",{"name":201,"orcid":9},"Joseph Chemutt",{"name":203,"orcid":9},"Ange Marie Tesire",{"name":205,"orcid":9},"Samuel Ishimwe",{"name":207,"orcid":208},"Yves Hategekimana","https:\u002F\u002Forcid.org\u002F0000-0002-2430-8072",{"name":210,"orcid":9},"Lilian Ndungu",{"name":212,"orcid":9},"Golo Rademacher",{"name":214,"orcid":215},"Aniruddha Ghosh","https:\u002F\u002Forcid.org\u002F0000-0003-3667-8019",{"tldr":217,"method":218,"finding":219,"direction":104,"opportunity":220},"在卢旺达四区采集四种主粮作物的GPS标注地块数据集，支撑GeoAI作物监测。","KoboToolbox结合JECAM指南与FAO分类，用Planet影像数字化9","获得2739个GPS点与914个作物标注多边形，含农艺管理信息。","可基于该数据集开发小样本\u002F迁移学习作物制图模型，并扩展至更多撒哈拉以南非洲国家。","2026-10-03T23:30:35.255650Z",{"id":223,"title":224,"url":225,"summary":226,"summary_zh":227,"content":9,"source_name":228,"source_url":225,"published_at":229,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":230,"score_detail":231,"sources":234,"tags":236,"primary_topic":239,"secondary_topics":240,"classification":245,"search_phrases":261,"slug":264,"view_count":98,"doi":265,"paper":266,"created_at":281},5592,"Evaluation of Concept-Rank Collapse Audits for Computer Vision Agriculture Disease Classifiers Under Low-Label Regimes: An Interpretability Capacity Bound for Smallholder Fungicide Selectivity","https:\u002F\u002Fdoi.org\u002F10.5281\u002Fzenodo.23545193","Concept-based interpretability methods for plant-disease classifiers are typically evaluated by whether their concept activations correlate with human-annotated labels, but this criterion is silent on a prior question: whether the classifier's internal representation retains enough concept-rank structure for any audit to be informative. the present analysis formalises that prior question as an interpretability capacity bound. A classifier is modelled as a concept-rank operator whose singular values govern how much concept-level signal survives the projection from pixel space to class logits, and a collapse audit is defined as the ratio of retained concept-rank energy to total concept energy. Under a low-label regime, where concept annotations are scarce and the audit must be estimated from few labelled examples, the bound predicts that audit reliability degrades as a function of the concept-rank spectrum and the label budget. The analysis is a theoretical and methodological contribution: it derives the bound, states the conditions under which a collapse audit is identifiable, and characterises the operating characteristics of the audit as a decision procedure under a declared variance structure (between-cluster SD = 6.0, residual SD = 4.5, intraclass correlation = 0.64). Power was estimated across 2000 design cells for cluster counts of 8, 16, 24 and 32 and effect sizes of 3, 5 and 8 percentage points; 52 clusters were required for 5 percentage points at 80% power. The bound reframes concept audits as capacity tests rather than fidelity tests, with consequences for how smallholder fungicide-selectivity decisions should be supported by interpretable disease classifiers.","面向植物病害分类器的基于概念的可解释性方法，通常以其概念激活是否与人工标注标签相关作为评价标准，但该标准对一个更先前的问题保持沉默：分类器的内部表征是否保留了足够的概念秩结构，从而使任何审计具有信息量。本分析将该先前问题形式化为一个可解释性容量上界。分类器被建模为概念秩算子，其奇异值支配着从像素空间到类别logits的投影中概念级信号得以保留的程度，并将坍缩审计定义为保留的概念秩能量与总概念能量之比。在低标签情形下，概念标注稀缺且审计必须由少量标注样本估计，该上界预测审计可靠性会随概念秩谱和标签预算的函数而退化。本分析是一项理论和方法论贡献：推导出该上界，陈述坍缩审计可识别的条件，并在声明的方差结构（簇间SD = 6.0，残差SD = 4.5，组内相关系数 = 0.64）下刻画审计作为决策程序的操作特征。在2000个设计单元中，针对簇数为8、16、24和32以及效应量为3、5和8个百分点的情形估计了功效；在80%功效下达到5个百分点需要52个簇。该上界将概念审计重新框定为容量检验而非保真度检验，并对可解释病害分类器应如何支持小农杀菌剂选择性决策具有启示意义。","Zenodo (CERN European Organization for Nuclear Research)","2026-10-21T00:00:00Z",69,{"impact":20,"substance":232,"depth":17,"authority":51,"freshness":13,"relevant":21,"comment":233},21,"理论方法论文，提出可解释性容量上界并给出审计功效分析，对智慧农业病害识别可信度有参考价值，但属小众方法学研究、缺乏田间实证，暂不宜进入每日精选。",[235],{"name":228,"url":225},[26,27,29,237,238],"植物病害识别","可解释性",{"code":32,"slug":33,"name":34,"path":35,"parent_code":36},[241],{"code":76,"slug":242,"name":243,"path":244,"parent_code":43},"agricultural-ai","农业AI与大模型","\u002Ftopic\u002Fdigital-agriculture\u002Fagricultural-ai",{"status":45,"locked":14,"version":46,"method":47,"evidence":246},{"candidates":247,"agricultural_evidence":52},[248,252,257],{"code":32,"score":51,"supported":52,"title":249,"abstract":250,"tags":251},[54],[54],[29],{"code":76,"score":13,"supported":52,"title":253,"abstract":255,"tags":256},[254],"computer vision",[],[27,237],{"code":71,"score":21,"supported":14,"title":258,"abstract":259,"tags":260},[],[],[26],[262,263],"农业人工智能 植物病害识别 可解释性 智慧农业","农业人工智能 植物病害识别","农业人工智能植物病害识别可解释性智慧农业-5592","10.5281\u002Fzenodo.23545193",{"doi":265,"openalex_id":267,"authors":268,"venue":228,"cited_by_count":98,"oa_url":225,"card":275,"direction":279,"ingested_from":107},"W7223137323",[269,271,273],{"name":270,"orcid":9},"Akshay Ramgoolam",{"name":272,"orcid":9},"Priya Bundhoo Appadoo",{"name":274,"orcid":9},"Jean-Noël LEspérance",{"tldr":276,"method":277,"finding":278,"direction":279,"opportunity":280},"提出可解释性容量上界，评估低标签下概念秩崩溃审计的可靠性。","将分类器建模为概念秩算子，推导容量上界并做2000格功效模拟。","审计可靠性随概念秩谱与标签预算下降；5个百分点效应需52个聚类达80%功效。","农业人工智能与决策模型","可将容量上界用于小农户杀菌剂选择性决策的可解释模型审计与标签预算设计。","2026-10-22T11:21:10.425290Z",{"id":283,"title":284,"url":285,"summary":286,"summary_zh":287,"content":9,"source_name":228,"source_url":285,"published_at":229,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":288,"score_detail":289,"sources":291,"tags":293,"primary_topic":296,"secondary_topics":297,"classification":299,"search_phrases":315,"slug":318,"view_count":98,"doi":319,"paper":320,"created_at":334},5590,"Evaluation of Concept-Probe Temporal-Stability Audits for Computer Vision Agriculture Disease Classifiers Under Seasonal Phenology Drift: An Interpretability Decay Bound for Smallholder Fungicide Timing","https:\u002F\u002Fdoi.org\u002F10.5281\u002Fzenodo.23545234","Concept-probe audits are increasingly used to certify that a crop-disease classifier's internal representations track the pathological concepts a smallholder would act on, but a probe certified at one point in the season may silently decouple from those concepts as phenology shifts the visual statistics of the field. the present analysis treats that decoupling as a formal object. It defines a temporal-stability audit as a sequence of concept-probe accuracies measured across defined phenological windows, and it derives an interpretability decay bound: the rate at which probe accuracy can fall before the fungicide-timing decision it is meant to support becomes unreliable. The bound is expressed as a function of the probe's initial margin, the drift in concept prevalence between windows, and the classifier's sensitivity to the drifted features, and it yields a certified audit interval rather than a single-point certificate. The paper then characterises the operating characteristics of the audit design under a declared variance structure, using a replication-based evaluation of the before-after-control-impact estimator with between-cluster SD = 6.0, residual SD = 4.5 and intraclass correlation = 0.64. Power was estimated across 2000 design cells over cluster counts of 8, 16, 24 and 32 and effect sizes of 3, 5 and 8 percentage points; 54 clusters were required for 5 percentage points at 80% power. The contribution is a decision-theoretic account of when a concept-probe certificate expires, and a design rule for scheduling re-audits across the season.","概念探针审计（concept-probe audit）日益被用于验证作物病害分类器的内部表征是否追踪小农户会据以采取行动的病理学概念，但在生长季某一时点获得认证的探针，可能随着物候变化改变田间的视觉统计特征而悄然与这些概念脱钩。本分析将这种脱钩视为一个形式化对象。它将时间稳定性审计（temporal-stability audit）定义为在既定物候窗口内测量的一系列概念探针准确率，并推导出一个可解释性衰减界限（interpretability decay bound）：即探针准确率在其所支持的杀菌剂施用时机决策变得不可靠之前所能下降的速率。该界限表示为探针初始裕度、窗口间概念流行率漂移以及分类器对漂移特征敏感度的函数，并给出一个经认证的审计区间而非单点认证。随后，本文在声明的方差结构下刻画该审计设计的运行特征，采用基于重复的前后对照影响估计量评估，其中组间标准差 = 6.0、残差标准差 = 4.5、组内相关系数 = 0.64。功效在 2000 个设计单元上估计，聚类数取 8、16、24 和 32，效应量取 3、5 和 8 个百分点；在 80% 功效下达到 5 个百分点需要 54 个聚类。本文的贡献在于从决策论角度阐明概念探针认证何时失效，并给出在整个生长季内安排重新审计的设计规则。",70,{"impact":20,"substance":18,"depth":17,"authority":51,"freshness":13,"relevant":21,"comment":290},"该论文提出概念探针审计的时序稳定性框架与可解释性衰减界，方法新颖、统计设计扎实，但属小众方法论研究，产业影响有限，适合作为专业精选而非头条。",[292],{"name":228,"url":285},[26,27,29,294,295],"作物病害识别","模型审计",{"code":32,"slug":33,"name":34,"path":35,"parent_code":36},[298],{"code":76,"slug":242,"name":243,"path":244,"parent_code":43},{"status":45,"locked":14,"version":46,"method":47,"evidence":300},{"candidates":301,"agricultural_evidence":52},[302,307,311],{"code":32,"score":303,"supported":52,"title":304,"abstract":305,"tags":306},15,[54],[29,54],[29],{"code":76,"score":13,"supported":52,"title":308,"abstract":309,"tags":310},[254],[],[294,27],{"code":71,"score":21,"supported":14,"title":312,"abstract":313,"tags":314},[],[],[26],[316,317],"概念探针 时序稳定性 作物病害分类器","杀菌剂 施药时机 可解释性衰减","概念探针时序稳定性作物病害分类器-5590","10.5281\u002Fzenodo.23545234",{"doi":319,"openalex_id":321,"authors":322,"venue":228,"cited_by_count":98,"oa_url":285,"card":329,"direction":279,"ingested_from":107},"W7223088030",[323,325,327],{"name":324,"orcid":9},"Abdi-Basid Idriss Farah",{"name":326,"orcid":9},"Amina Abdillahi Guedi",{"name":328,"orcid":9},"Moussa Omar Youssouf",{"tldr":330,"method":331,"finding":332,"direction":279,"opportunity":333},"提出概念探针时间稳定性审计与可解释性衰减界，判断何时需重新审计。","定义跨物候窗口探针精度序列，推导衰减界，用BACI复制评估统计功效。","5个百分点效应需54个聚类达80%功效，审计证书会随物候漂移过期。","可将衰减界与再审计调度规则嵌入小农户杀菌剂时机决策支持系统，并验证跨作物泛化。","2026-10-22T11:21:08.978348Z",{"id":336,"title":337,"url":338,"summary":339,"summary_zh":340,"content":9,"source_name":341,"source_url":338,"published_at":229,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":342,"score_detail":343,"sources":345,"tags":347,"primary_topic":350,"secondary_topics":356,"classification":364,"search_phrases":396,"slug":399,"view_count":98,"doi":400,"paper":401,"created_at":427},5573,"Improving alfalfa nitrogen estimation with canopy structural priors from DSSAT–PROSAIL and UAV hyperspectral data","https:\u002F\u002Fdoi.org\u002F10.1016\u002Fj.eja.2026.128375","Accurate estimation of nitrogen (N) concentration in alfalfa ( Medicago sativa L .) supports forage quality assessment and harvest scheduling. Empirical remote sensing models often overfit because of confounding effects of three-dimensional canopy structure. Incorporating structural parameters can stabilize biochemical inversions, but concurrent in situ morphological measurements are difficult to acquire at scale. We developed the Morphology-Aware Latent Embedding Network (MALENet), a hybrid framework integrating physiological and morphological mechanisms with hyperspectral data. The Decision Support System for Agrotechnology Transfer (DSSAT), coupled with the PROSAIL radiative transfer model, generated a structurally rich synthetic hyperspectral prior. A convolutional neural network (CNN) with a Squeeze-and-Excitation (SE) attention mechanism, was pre-trained on these data to distill ecophysiological knowledge into a dense latent representation. This mechanistically derived embedding was then fused with empirical hyperspectral measurements collected under diverse field conditions. Evaluation using multi-year field quadrats in the Hexi Corridor showed improved predictive performance ( R² = 0.71, RMSE = 0.30%) relative to conventional machine learning and deep learning models (LASSO-ANN: R² = 0.61; CNN: R² = 0.66). Feature attribution indicated that MALENet used near-infrared multiple scattering (58.3% attribution) as a structural proxy to calibrate nitrogen concentration retrieval against the confounding effects of canopy architecture. Despite constraints imposed by sample size and PROSAIL's 1D canopy assumption, these findings demonstrate the potential of physically guided deep learning to use structural information for agricultural trait retrieval in precision forage management.","准确估算紫花苜蓿（Medicago sativa L.）的氮浓度有助于饲草品质评估和收获期安排。经验性遥感模型常因三维冠层结构的混杂效应而出现过拟合。引入结构参数可以稳定生化反演，但同步开展原位形态学测量难以大规模获取。我们开发了形态感知潜在嵌入网络（Morphology-Aware Latent Embedding Network，MALENet），这是一个将生理与形态机制同高光谱数据相结合的混合框架。农业技术转移决策支持系统（Decision Support System for Agrotechnology Transfer，DSSAT）与PROSAIL辐射传输模型耦合，生成了结构信息丰富的合成高光谱先验数据。在此基础上，预训练了一个带有挤压—激励（Squeeze-and-Excitation，SE）注意力机制的卷积神经网络（convolutional neural network，CNN），将生态生理学知识提炼为稠密潜在表征。随后，将这一机制性导出的嵌入与在不同田间条件下采集的经验性高光谱测量数据相融合。利用河西走廊多年田间样方进行的评估显示，相较于传统机器学习和深度学习模型（LASSO-ANN：R² = 0.61；CNN：R² = 0.66），预测性能有所提升（R² = 0.71，RMSE = 0.30%）。特征归因表明，MALENet利用近红外多次散射（归因占比58.3%）作为结构代理，以校准氮浓度反演，抵御冠层结构的混杂效应。尽管受样本量和PROSAIL一维冠层假设的限制，这些发现表明，物理引导的深度学习有潜力利用结构信息，在精准饲草管理中进行农业性状反演。","European Journal of Agronomy",80,{"impact":122,"substance":18,"depth":17,"authority":123,"freshness":13,"relevant":21,"comment":344},"该研究将作物模型与辐射传输模型耦合生成结构先验，并结合无人机高光谱数据反演苜蓿氮素，方法新颖、田间验证扎实，对精准饲草管理有参考价值，值得进入每日精选。",[346],{"name":341,"url":338},[26,27,28,348,349],"氮素监测","紫花苜蓿",{"code":351,"slug":352,"name":353,"path":354,"parent_code":355},"A7","forage-cropping","饲草与耕作制度","\u002Ftopic\u002Ffield-crops\u002Fforage-cropping","A",[357,358],{"code":39,"slug":40,"name":41,"path":42,"parent_code":43},{"code":359,"slug":360,"name":361,"path":362,"parent_code":363},"C4","feed","饲料与饲草","\u002Ftopic\u002Flivestock\u002Ffeed","C",{"status":45,"locked":14,"version":46,"method":47,"evidence":365},{"candidates":366,"agricultural_evidence":52},[367,372,378,384,392],{"code":39,"score":17,"supported":52,"title":368,"abstract":370,"tags":371},[369],"hyperspectral",[60,61,369],[28],{"code":351,"score":303,"supported":52,"title":373,"abstract":375,"tags":377},[374],"alfalfa",[376,374],"苜蓿",[349],{"code":359,"score":13,"supported":52,"title":379,"abstract":380,"tags":383},[],[381,376,382],"饲草","forage",[349],{"code":76,"score":13,"supported":52,"title":385,"abstract":386,"tags":391},[],[387,388,389,390,148],"深度学习","机器学习","machine learning","deep learning",[27],{"code":71,"score":21,"supported":14,"title":393,"abstract":394,"tags":395},[],[],[26],[397,398],"DSSAT PROSAIL 苜蓿 氮素","无人机 高光谱 苜蓿","DSSATPROSAIL苜蓿氮素-5573","10.1016\u002Fj.eja.2026.128375",{"doi":400,"openalex_id":402,"authors":403,"venue":341,"cited_by_count":98,"oa_url":338,"card":422,"direction":104,"ingested_from":107},"W7223154576",[404,406,408,410,412,415,417,420],{"name":405,"orcid":9},"Shuai Fu",{"name":407,"orcid":9},"Yunhao Li",{"name":409,"orcid":9},"Chunli Miao",{"name":411,"orcid":9},"Jie Liu",{"name":413,"orcid":414},"Tiangang Liang","https:\u002F\u002Forcid.org\u002F0000-0001-5599-6587",{"name":416,"orcid":9},"Yongjia Hu",{"name":418,"orcid":419},"Qisheng Feng","https:\u002F\u002Forcid.org\u002F0009-0000-1391-9591",{"name":421,"orcid":9},"吴彩霞",{"tldr":423,"method":424,"finding":425,"direction":104,"opportunity":426},"构建MALENet融合DSSAT-PROSAIL结构先验与无人机高光谱，提升苜蓿氮浓度估算精度。","DSSAT耦合PROSAIL生成合成高光谱先验，预训练CNN-SE网络，再融合田","MALENet精度优于传统机器学习和深度学习模型，R²达0.71，近红外多次散射作为结构代理贡献58","可探索将作物模型结构先验迁移至其他牧草或作物生化参数反演，并缓解PROSAIL一维冠层假设局限。","2026-10-22T11:20:39.622844Z",{"id":429,"title":430,"url":431,"summary":432,"summary_zh":433,"content":9,"source_name":434,"source_url":431,"published_at":435,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":342,"score_detail":436,"sources":439,"tags":441,"primary_topic":444,"secondary_topics":445,"classification":446,"search_phrases":475,"slug":478,"view_count":21,"doi":479,"paper":480,"created_at":497},4563,"Seeing Crops in 3D: Remote Sensing for Smarter Precision Farming","https:\u002F\u002Fdoi.org\u002F10.12688\u002Fopenreseurope.25171.1","The growing adoption of digital technologies in agriculture is opening new opportunities for improving productivity, sustainability and resource efficiency through data-driven decision making. This paper presents the integrated precision-agriculture framework developed by RFSAT Limited within the H2020 AgriBIT project and currently being further enhanced through the ENACT project. The framework combines Internet-of-Things (IoT) sensing, centimetre-accuracy Global Navigation Satellite System (GNSS) positioning, high-resolution multispectral remote sensing and Artificial Intelligence (AI) analytics into a single smart-farming solution. A key enabler is a low-cost, centimetre-accurate GNSS receiver developed by RFSAT, which allows precise georeferencing of aerial and ground observations and, in turn, the generation of high-resolution multispectral 3D models of crop canopies with spatial resolution exceeding 1 cm\u002Fpixel. These digital representations support detailed monitoring of crop growth, canopy development, biomass distribution and plant health throughout the season. Fused with an IoT monitoring infrastructure of environmental and microclimatic sensors, the georeferenced remote-sensing products form a comprehensive digital twin of the agricultural environment, over which machine-learning techniques automatically identify and localise crop anomalies including water stress, nutrient deficiencies, disease outbreaks and pest infestations. The technologies have been validated at operational pilot sites across Greece, Portugal and Italy, spanning orchards, vineyards and other high-value crops, where the combination of centimetre positioning, multispectral 3D modelling, IoT monitoring and AI analytics substantially improved monitoring accuracy, reduced input consumption and enhanced management decisions.","数字技术在农业中的日益普及，正通过数据驱动的决策为提升生产力、可持续性和资源效率开辟新的机遇。本文介绍了RFSAT有限公司在H2020 AgriBIT项目中开发的综合精准农业框架，该框架目前正通过ENACT项目进一步完善。该框架将物联网（IoT）传感、厘米级精度的全球导航卫星系统（GNSS）定位、高分辨率多光谱遥感以及人工智能（AI）分析整合为一个智慧农业解决方案。一个关键的使能因素是由RFSAT开发的低成本、厘米级精度GNSS接收器，它能够对航空和地面观测进行精确的地理参照，进而生成空间分辨率超过1厘米\u002F像素的作物冠层高分辨率多光谱三维模型。这些数字表示支持在整个生长季对作物生长、冠层发育、生物量分布和植物健康进行详细监测。通过与由环境和小气候传感器组成的物联网监测基础设施相融合，地理参照的遥感产品形成了农业环境的综合数字孪生，机器学习技术在此基础上自动识别和定位作物异常，包括水分胁迫、养分缺乏、病害暴发和虫害侵扰。这些技术已在希腊、葡萄牙和意大利的运营试点基地进行了验证，涵盖果园、葡萄园和其他高价值作物，其中厘米级定位、多光谱三维建模、物联网监测和AI分析的结合大幅提高了监测精度、降低了投入品消耗并改善了管理决策。","Open Research Europe","2026-10-09T00:00:00Z",{"impact":17,"substance":18,"depth":437,"authority":19,"freshness":20,"relevant":21,"comment":438},19,"欧盟H2020项目论文，融合厘米级GNSS、多光谱3D建模与AI分析，方法新颖且多国验证，对智慧农业技术路线有参考价值。",[440],{"name":434,"url":431},[26,27,28,442,443],"数字孪生","精准农业",{"code":39,"slug":40,"name":41,"path":42,"parent_code":43},[],{"status":45,"locked":14,"version":46,"method":47,"evidence":447},{"candidates":448,"agricultural_evidence":52},[449,454,459,463,470],{"code":39,"score":17,"supported":52,"title":450,"abstract":451,"tags":453},[61],[60,61,452],"multispectral",[28],{"code":71,"score":51,"supported":52,"title":455,"abstract":456,"tags":458},[],[26,443,442,457],"digital twin",[442,26,443],{"code":76,"score":144,"supported":52,"title":460,"abstract":461,"tags":462},[],[388,147],[27],{"code":464,"score":110,"supported":52,"title":465,"abstract":466,"tags":469},"I5",[],[467,468],"物联网","IoT",[],{"code":179,"score":65,"supported":52,"title":471,"abstract":472,"tags":474},[],[473],"葡萄",[],[476,477],"AgriBIT 精准农业 遥感","GNSS 多光谱 3D 作物","AgriBIT精准农业遥感-4563","10.12688\u002Fopenreseurope.25171.1",{"doi":479,"openalex_id":481,"authors":482,"venue":434,"cited_by_count":98,"oa_url":431,"card":492,"direction":106,"ingested_from":107},"W7221178461",[483,486,489],{"name":484,"orcid":485},"Artur Krukowski","https:\u002F\u002Forcid.org\u002F0000-0002-6635-7585",{"name":487,"orcid":488},"Emmanouela Vogiatzaki","https:\u002F\u002Forcid.org\u002F0000-0001-9239-2045",{"name":490,"orcid":491},"İzzet Kale","https:\u002F\u002Forcid.org\u002F0000-0001-5562-6885",{"tldr":493,"method":494,"finding":495,"direction":104,"opportunity":496},"提出融合物联网、厘米级GNSS、多光谱遥感和AI的精准农业框架，构建作物数字孪生。","低成本厘米级GNSS、多光谱3D建模、IoT传感器与机器学习，在希腊、葡萄牙、意","厘米定位与多光谱3D建模结合AI可提升监测精度、减少投入并改善管理决策。","可探索低成本厘米级GNSS与多光谱3D数字孪生在更多作物和区域的可扩展性及实时决策优化。","2026-10-10T23:30:07.913557Z",{"id":499,"title":500,"url":501,"summary":502,"summary_zh":503,"content":9,"source_name":504,"source_url":501,"published_at":505,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":506,"score_detail":507,"sources":511,"tags":513,"primary_topic":514,"secondary_topics":516,"classification":518,"search_phrases":541,"slug":544,"view_count":545,"doi":546,"paper":547,"created_at":575},4356,"Dual‐Branch Network With Enhanced Feature Fusion for Improved Rice Classification Accuracy","https:\u002F\u002Fdoi.org\u002F10.1002\u002Fsae2.70218","ABSTRACT Accurate mapping of rice, one of the world's most important food crops, is critical for monitoring its growth, estimating yields, and supporting global food security. Most deep learning methods for crop classification rely on multi‐temporal optical remote sensing data. However, it is difficult to obtain sufficient ground‐truth samples for time‐series deep learning models through conducting large‐scale field surveys, and persistent cloud and rainfall often limit optical observations during the rice‐growing seasons in subtropical regions. Rice classification based solely on time‐series optical images may therefore be less accurate and reliable. Here, we developed two dual‐branch deep learning networks for effective rice identification: one combining a long short‐term memory network running in parallel with a multi‐layer perceptron (LSTM + MLP), and the other replacing the LSTM with a one‐dimensional convolutional neural network (Conv1d + MLP). These models process temporal sequences in parallel with auxiliary data through the MLP, allowing them to automatically extract phenology information while incorporating auxiliary variables such as texture features, digital elevation model (DEM) data, and the standard deviation of the vegetation index time series. We evaluated these models in a representative county in the subtropical region of China, and compared the accuracies with standalone LSTM, Conv1d, MLP, and random forest (RF) models. The results revealed that LSTM + MLP had the highest accuracy, with an F1‐score of 95.5%, which can automatically extract rice growth patterns and successfully identify the key phenological information of flood signals during the rice transplanting period. By incorporating auxiliary variables, the model achieved an average accuracy improvement of 8.1% relative to the models using only temporal features. The proposed method provides a useful approach for rice mapping in subtropical regions.","准确绘制水稻这一全球最重要的粮食作物之一，对于监测其生长、估算产量以及保障全球粮食安全至关重要。大多数作物分类的深度学习方法依赖于多时相光学遥感数据。然而，通过开展大规模实地调查来获取足够的地面真值样本以用于时间序列深度学习模型十分困难，而且亚热带地区水稻生长季期间持续的云层和降雨往往限制了光学观测。因此，仅基于时间序列光学图像的水稻分类可能精度较低且可靠性不足。本研究开发了两种双分支深度学习网络以实现有效的水稻识别：一种将长短期记忆网络与多层感知机并行运行（LSTM + MLP），另一种则用一维卷积神经网络替代LSTM（Conv1d + MLP）。这些模型通过MLP将时间序列与辅助数据并行处理，使其能够自动提取物候信息，同时纳入纹理特征、数字高程模型（DEM）数据以及植被指数时间序列标准差等辅助变量。我们在中国亚热带地区一个代表性县对这些模型进行了评估，并将精度与单独的LSTM、Conv1d、MLP和随机森林（RF）模型进行了比较。结果表明，LSTM + MLP精度最高，F1分数达到95.5%，能够自动提取水稻生长模式，并成功识别水稻移栽期洪水信号的关键物候信息。通过纳入辅助变量，该模型相对于仅使用时间特征的模型平均精度提高了8.1%。所提出的方法为亚热带地区水稻制图提供了一种有效途径。","Journal of Sustainable Agriculture and Environment","2026-10-06T00:00:00Z",75,{"impact":303,"substance":232,"depth":508,"authority":19,"freshness":509,"relevant":21,"comment":510},17,9,"提出双分支深度学习网络融合时序与辅助数据，水稻识别F1达95.5%，方法新颖且结论可靠，对亚热带多云雨区作物遥感制图有实用价值。",[512],{"name":504,"url":501},[26,27,154,60,30],{"code":151,"slug":155,"name":154,"path":515,"parent_code":355},"\u002Ftopic\u002Ffield-crops\u002Frice",[517],{"code":39,"slug":40,"name":41,"path":42,"parent_code":43},{"status":45,"locked":14,"version":46,"method":47,"evidence":519},{"candidates":520,"agricultural_evidence":52},[521,525,529,533,537],{"code":151,"score":303,"supported":52,"title":522,"abstract":523,"tags":524},[155],[154,155],[154],{"code":39,"score":20,"supported":52,"title":526,"abstract":527,"tags":528},[],[60,61],[30,60],{"code":76,"score":144,"supported":52,"title":530,"abstract":531,"tags":532},[],[387,390],[27],{"code":172,"score":110,"supported":52,"title":534,"abstract":535,"tags":536},[],[175,176],[],{"code":71,"score":21,"supported":14,"title":538,"abstract":539,"tags":540},[],[],[26],[542,543],"水稻 遥感 分类 深度学习","亚热带 水稻 制图","水稻遥感分类深度学习-4356",4,"10.1002\u002Fsae2.70218",{"doi":546,"openalex_id":548,"authors":549,"venue":504,"cited_by_count":98,"oa_url":569,"card":570,"direction":104,"ingested_from":107},"W7220691059",[550,552,555,558,560,563,566],{"name":551,"orcid":9},"Rongfei Ma",{"name":553,"orcid":554},"Wenjiao Shi","https:\u002F\u002Forcid.org\u002F0000-0001-6199-9733",{"name":556,"orcid":557},"Minglei Wang","https:\u002F\u002Forcid.org\u002F0000-0002-4174-9224",{"name":559,"orcid":9},"Xiwei Chen",{"name":561,"orcid":562},"Chenchen Zhang","https:\u002F\u002Forcid.org\u002F0000-0003-0458-6232",{"name":564,"orcid":565},"Baihong Pan","https:\u002F\u002Forcid.org\u002F0009-0004-1100-6798",{"name":567,"orcid":568},"Xiangming Xiao","https:\u002F\u002Forcid.org\u002F0000-0003-0956-7428","https:\u002F\u002Fonlinelibrary.wiley.com\u002Fdoi\u002Fpdfdirect\u002F10.1002\u002Fsae2.70218",{"tldr":571,"method":572,"finding":573,"direction":104,"opportunity":574},"提出双分支深度学习网络融合时序与辅助数据，提升亚热带水稻分类精度。","LSTM\u002FConv1d与MLP并行双分支网络，融合时序植被指数、纹理、DEM等辅","LSTM+MLP模型F1达95.5%，辅助变量使精度平均提升8.1%，能识别插秧期淹水信号。","可探索多源遥感与气象数据融合，解决多云雨地区时序缺失下的作物精细制图问题。","2026-10-07T23:30:25.945309Z"]