[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"item-1748":3,"related-1748":84},{"id":4,"title":5,"url":6,"summary":7,"summary_zh":8,"source_name":9,"source_url":6,"published_at":10,"category":11,"cover_url":12,"hotness":13,"is_selected":14,"embedding_ready":15,"score":16,"sources":17,"tags":19,"primary_topic":21,"secondary_topics":27,"search_phrases":28,"slug":29,"view_count":30,"doi":31,"paper":32,"created_at":52,"content":12,"score_detail":53,"classification":55},1748,"Cross-Layer Energy Management in Embedded AIoT Sensor Networks: A Structured Review, Full-Cycle Energy Model, and Break-Even Analysis","https:\u002F\u002Fdoi.org\u002F10.3390\u002Felectronics15174000","Energy management in battery-powered embedded Artificial Intelligence of Things (AIoT) nodes requires coordinated control of sensing, computation, and wireless communication. This structured review combines a documented search of Scopus, Web of Science Core Collection, and IEEE Xplore with screening of 548 unique records and full-text source-to-construct coding of 56 database-derived sources. The synthesis shows that acquisition, Tiny Machine Learning (TinyML) inference, communication, protocol-state overhead, adaptive sensing, and energy harvesting are commonly optimized separately, while field-validated joint control remains limited. The review develops a full-cycle energy model and cross-layer framework linking sensing, processing, payload representation, radio policy, and energy availability. Dimensionless break-even analysis shows that the admissible local-processing budget is bounded by fixed communication overhead and non-machine-learning edge costs. A targeted Long-Term Evolution for Machines (LTE-M) experiment comprising 40 complete transactions showed that payload length alone was insufficient to characterize communication energy; a state-aware model incorporating measured signaling duration achieved R2=0.980 and predicted a temporally separated second session with 1.66% relative root mean square error (RMSE). Precision agriculture and precision apiculture illustrate modality- and network-dependent trade-offs. The findings support full-cycle benchmarking and context-aware orchestration rather than isolated optimization of inference or communication.","电池供电的嵌入式人工智能物联网（AIoT）节点中的能量管理需要对感知、计算和无线通信进行协调控制。本结构化综述结合了对Scopus、Web of Science核心合集及IEEE Xplore的文献检索，筛选了548条独特记录，并对56篇数据库来源文献进行了全文源到构念编码。综合结果表明，数据采集、微型机器学习（TinyML）推理、通信、协议状态开销、自适应感知和能量采集通常被分别优化，而经过现场验证的联合控制仍然有限。本综述构建了一个全周期能量模型及跨层框架，将感知、处理、载荷表示、无线电策略和能量可用性联系起来。无量纲盈亏平衡分析表明，可允许的本地处理预算受固定通信开销和非机器学习边缘成本的约束。一项针对机器类通信长期演进（LTE-M）的定向实验包含40次完整事务，结果显示仅凭载荷长度不足以表征通信能量；一种纳入实测信令持续时间的状态感知模型达到了R²=0.980，并以1.66%的相对均方根误差（RMSE）预测了时间上分离的第二个会话。精准农业和精准养蜂展示了依赖模态和网络的权衡关系。研究结果支持全周期基准测试和情境感知编排，而非对推理或通信的孤立优化。","Electronics","2026-09-04T00:00:00Z","论文",null,10,false,true,0,[18],{"name":9,"url":6},[20],"智慧农业",{"code":22,"slug":23,"name":24,"path":25,"parent_code":26},"I2","agricultural-ai","农业AI与大模型","\u002Ftopic\u002Fdigital-agriculture\u002Fagricultural-ai","I",[],[],"Cross-LayerEnergyManagementinEmbeddedAIo-1748",1,"10.3390\u002Felectronics15174000",{"doi":31,"openalex_id":33,"authors":34,"venue":9,"cited_by_count":16,"oa_url":6,"card":44,"direction":50,"ingested_from":51},"W7208717295",[35,38,41],{"name":36,"orcid":37},"Sebastian Górecki","https:\u002F\u002Forcid.org\u002F0000-0001-5700-4000",{"name":39,"orcid":40},"Sławomir Hausman","https:\u002F\u002Forcid.org\u002F0000-0003-3891-4214",{"name":42,"orcid":43},"Piotr Korbel","https:\u002F\u002Forcid.org\u002F0000-0003-3716-0932",{"tldr":45,"method":46,"finding":47,"direction":48,"opportunity":49},"综述嵌入式AIoT传感器网络的跨层能量管理，提出全周期能量模型和盈亏平衡分析。","结构化综述、全周期能量模型、无量纲盈亏平衡分析及LTE-M实验验证。","局部处理预算受通信开销限制，状态感知模型预测通信能耗准确（R²=0.980）。","其他","可探索农业场景下跨层联合优化，如结合作物生长模型动态调整感知与通信策略。","农业人工智能与决策模型","openalex","2026-09-05T23:30:45.356563Z",{"impact":16,"substance":16,"depth":16,"authority":16,"freshness":16,"relevant":16,"comment":54},"该论文聚焦嵌入式AIoT传感器网络的能源管理，虽提及精准农业，但核心主题并非农业信息化，相关性不足。",{"status":56,"locked":14,"version":57,"method":58,"evidence":59},"classified","2026-10-22.2","rules",{"candidates":60,"agricultural_evidence":15},[61,69,77],{"code":22,"score":62,"supported":15,"title":63,"abstract":64,"tags":68},9,[],[65,66,67],"机器学习","machine learning","artificial intelligence",[],{"code":70,"score":71,"supported":15,"title":72,"abstract":73,"tags":76},"I1",7,[],[74,75],"精准农业","precision agriculture",[20],{"code":78,"score":79,"supported":15,"title":80,"abstract":81,"tags":83},"I5",3,[],[82],"物联网",[],{"total":85,"page":30,"page_size":85,"items":86},6,[87,181,221,260,313,355],{"id":88,"title":89,"url":90,"summary":91,"summary_zh":92,"source_name":93,"source_url":90,"published_at":94,"category":11,"cover_url":12,"hotness":13,"is_selected":14,"embedding_ready":15,"score":95,"sources":96,"tags":98,"primary_topic":103,"secondary_topics":108,"search_phrases":116,"slug":119,"view_count":16,"doi":120,"paper":121,"created_at":180},5614,"Replacing blanket agronomic recommendations with agroecology-specific decision support: a simulation-based framework for irrigated dry-season rice production in coastal deltas","https:\u002F\u002Fdoi.org\u002F10.1071\u002Fcp26175","Context Irrigated dry-season (Boro) rice (Oryza sativa L.) is a major contributor to food security in Bangladesh and inland West Bengal, but its cultivation is limited in the coastal Sundarbans due to high salinity, tidal influences, and scarcity of fresh irrigation water. Current policy frameworks promoting uniform transplanting windows and cultivar choices disregard the marked spatial heterogeneity of abiotic stress regimes in these deltaic agroecosystems. Aims This study aimed to characterize genotype × environment × management (G × E × M) interactions and identify location-specific sowing date–cultivar combinations to improve Boro rice productivity, irrigation water productivity, and yield stability under coastal saline conditions. Methods The Agricultural Production Systems Simulator (APSIM-Oryza) was calibrated and validated using multi-location field experimental data from Bangladesh and India collected during 2016–2018. Model performance was evaluated for phenology, above-ground biomass, and grain yield, with root mean square errors (RMSEs) of 6.7 days, 996 kg ha−1, and 1574 kg ha−1, respectively. Subsequently, 35 years of climate-driven simulations (1984–2018) were conducted to assess the effects of sowing dates and cultivar choices on crop performance, stress exposure, irrigation requirements, and irrigation water productivity across contrasting environments. Results Thirty-five years of climate-driven simulations (1984–2018) showed strong spatiotemporal variability in stress exposure. Early sowing (October) predisposed crops to cold-induced spikelet sterility, whereas late sowing (mid-December onwards) intensified exposure to terminal salinity and heat stress, elevating irrigation demand while reducing water productivity. Mid-November sowing consistently minimized risk trade-offs, achieving the highest probabilities of exceeding 5.0 Mg ha−1 with superior irrigation water productivity (9–11 kg grain ha−1 mm−1). Salt-tolerant cultivars (e.g. BRRI Dhan67, WGL20471) maintained yield across variable stress environments, whereas susceptible genotypes (e.g. BRRI Dhan28, IET4786) exhibited high interannual yield variability and frequent failure. Conclusions Sowing date and cultivar selection jointly determine Boro rice productivity and yield stability in coastal saline environments. Mid-November sowing, combined with salt-tolerant cultivars, emerged as a promising strategy for reducing climatic and salinity-related production risks. However, the magnitude of these benefits depends on local stress conditions and hydrological characteristics. Implications The findings support a shift from blanket regional recommendations towards dynamic, simulation-based decision-support systems that integrate sowing windows, stress-resilient cultivars, and location-specific hydrological and salinity conditions. Integrating these systems into agricultural extension services and policy frameworks could improve climate resilience, irrigation water productivity, and sustainable Boro rice production across the coastal saline zones of South Asia.","背景 灌溉旱季（Boro）水稻（Oryza sativa L.）是孟加拉国和西孟加拉邦内陆地区粮食安全的主要贡献者，但由于高盐度、潮汐影响和淡水灌溉水源稀缺，其种植在沿海孙德尔本斯地区受到限制。当前促进统一移栽窗口和品种选择的政策框架忽视了这些三角洲农业生态系统中非生物胁迫状况的显著空间异质性。目的 本研究旨在表征基因型×环境×管理（G×E×M）互作，并确定地点特定的播期—品种组合，以提高沿海盐渍条件下Boro水稻的生产力、灌溉水生产力和产量稳定性。方法 利用2016—2018年期间从孟加拉国和印度收集的多地点田间试验数据，对农业生产系统模拟器（APSIM-Oryza）进行了校准和验证。模型性能针对物候、地上生物量和籽粒产量进行了评估，均方根误差（RMSE）分别为6.7天、996 kg ha−1和1574 kg ha−1。随后，进行了35年气候驱动模拟（1984—2018），以评估播期和品种选择对 contrasting 环境下作物表现、胁迫暴露、灌溉需求和灌溉水生产力的影响。结果 35年气候驱动模拟（1984—2018）显示胁迫暴露存在强烈的时空变异性。早播（10月）使作物易受低温诱导的小穗不育影响，而晚播（12月中旬以后）则加剧了终期盐度和热胁迫的暴露，提高了灌溉需求同时降低了水分生产力。11月中旬播种持续最小化了风险权衡，实现了超过5.0 Mg ha−1的最高概率以及优越的灌溉水生产力（9—11 kg grain ha−1 mm−1）。耐盐品种（如BRRI Dhan67、WGL20471）在变化的胁迫环境中保持了产量，而敏感基因型（如BRRI Dhan28、IET4786）则表现出高的年际产量变异性和频繁的失败。结论 播期和品种选择共同决定了沿海盐渍环境中Boro水稻的生产力和产量稳定性。11月中旬播种结合耐盐品种，成为降低气候和盐度相关生产风险的有前景的策略。然而，其幅度","Australian Journal of Agricultural Research","2026-10-22T00:00:00Z",82,[97],{"name":93,"url":90},[20,99,100,101,102],"水稻","盐碱地","决策支持","灌溉管理",{"code":104,"slug":105,"name":99,"path":106,"parent_code":107},"A1","rice","\u002Ftopic\u002Ffield-crops\u002Frice","A",[109,115],{"code":110,"slug":111,"name":112,"path":113,"parent_code":114},"H6","biodiversity","生态保护与生物多样性","\u002Ftopic\u002Fenvironment\u002Fbiodiversity","H",{"code":22,"slug":23,"name":24,"path":25,"parent_code":26},[117,118],"APSIM-Oryza 水稻 模拟","沿海盐渍区 Boro 水稻 播期","APSIM-Oryza水稻模拟-5614","10.1071\u002Fcp26175",{"doi":120,"openalex_id":122,"authors":123,"venue":93,"cited_by_count":16,"oa_url":12,"card":175,"direction":50,"ingested_from":51},"W7223185956",[124,127,130,133,135,138,141,144,147,150,153,156,159,161,164,167,170,172],{"name":125,"orcid":126},"Sukamal Sarkar","https:\u002F\u002Forcid.org\u002F0000-0002-1438-1778",{"name":128,"orcid":129},"Donald S. Gaydon","https:\u002F\u002Forcid.org\u002F0000-0002-0078-4154",{"name":131,"orcid":132},"Mohammed Mainuddin","https:\u002F\u002Forcid.org\u002F0000-0002-6057-5688",{"name":134,"orcid":12},"Md. Maniruzzaman",{"name":136,"orcid":137},"M. B. Hossain","https:\u002F\u002Forcid.org\u002F0009-0003-5394-1381",{"name":139,"orcid":140},"Buddheswar Maji","https:\u002F\u002Forcid.org\u002F0000-0002-9924-6160",{"name":142,"orcid":143},"Sukanta Kumar Sarangi","https:\u002F\u002Forcid.org\u002F0000-0002-0356-7665",{"name":145,"orcid":146},"Anannya Dhar","https:\u002F\u002Forcid.org\u002F0009-0001-4160-2193",{"name":148,"orcid":149},"Saikat Dey","https:\u002F\u002Forcid.org\u002F0009-0000-7304-6219",{"name":151,"orcid":152},"Richard William Bell","https:\u002F\u002Forcid.org\u002F0000-0002-7756-3755",{"name":154,"orcid":155},"Mst Shetara Yesmin","https:\u002F\u002Forcid.org\u002F0009-0003-9249-8626",{"name":157,"orcid":158},"Palash Kumar Kundu","https:\u002F\u002Forcid.org\u002F0000-0001-6076-9493",{"name":160,"orcid":12},"Mark R. Glover",{"name":162,"orcid":163},"D. Burman","https:\u002F\u002Forcid.org\u002F0000-0002-2322-0044",{"name":165,"orcid":166},"Jatish Chandra Biswas","https:\u002F\u002Forcid.org\u002F0000-0002-4972-6676",{"name":168,"orcid":169},"Uttam Kumar Mandal","https:\u002F\u002Forcid.org\u002F0000-0002-7259-5037",{"name":171,"orcid":12},"K. K. Mahanta",{"name":173,"orcid":174},"E.G. Barrett-Lennard","https:\u002F\u002Forcid.org\u002F0000-0001-9945-1044",{"tldr":176,"method":177,"finding":178,"direction":50,"opportunity":179},"用APSIM模拟35年气候，为沿海盐渍稻区定制播期与品种决策。","APSIM-Oryza模型校准验证，1984–2018年多地点气候模拟。","11月中旬播种配耐盐品种可兼顾高产、节水与稳产。","可构建面向异质胁迫区的分区精准决策支持系统，替代统一农艺推荐。","2026-10-22T23:31:14.689445Z",{"id":182,"title":183,"url":184,"summary":185,"summary_zh":186,"source_name":187,"source_url":184,"published_at":94,"category":11,"cover_url":12,"hotness":13,"is_selected":14,"embedding_ready":15,"score":188,"sources":189,"tags":191,"primary_topic":195,"secondary_topics":196,"search_phrases":198,"slug":201,"view_count":16,"doi":202,"paper":203,"created_at":220},5613,"ExplainableCropAI: Interpretable Deep Learning for Agricultural Disease Diagnosis and Decision Support","https:\u002F\u002Fdoi.org\u002F10.1007\u002Fs44196-026-01654-9","Deep learning models classify crop diseases accurately, but their black-box nature limits practical use: farmers and extension officers need to understand a diagnosis before acting on it. We present ExplainableCropAI, a decision-support system for paddy disease diagnosis that attaches three complementary explanation channels to a ConvNeXt V2-Tiny classifier: Grad-CAM++ visual attribution, Monte Carlo Dropout uncertainty with temperature-scaled calibration, and natural-language explanations generated by Phi-3-mini under a grounding constraint that restricts the model to typed fields drawn from the prediction, the attribution map and a curated agronomic knowledge base. None of the components is new; the contribution lies in their integration into a single pipeline and in measuring the properties that integration is meant to deliver. On a leakage-controlled, held-out test split of 1,540 images from the Paddy Doctor dataset (10,407 images, nine diseases and healthy), the classifier reaches 97.5% accuracy (95% CI 96.7–98.3) and macro-F1 of 96.9%. Six attribution methods are compared under identical settings with model-based faithfulness metrics; the calibrated confidence separates correct from incorrect predictions with an AUROC of 0.93, and referring the 10% least-confident cases to an expert removes 74% of all errors. Under eight image corruptions at five severities the model retains 89.7% mean accuracy against 72–78% for six baselines trained identically, and its referral rate rises with the degree of degradation. Zero-shot evaluation on two independently collected rice-disease datasets from Bangladesh and India shows that accuracy roughly halves under cross-dataset shift and that confidence no longer identifies errors, while fifty local images per class recover accuracy above 90%. The code, split files, trained weights and evaluation scripts are released.","深度学习模型能够准确分类作物病害，但其黑箱性质限制了实际应用：农民和推广人员需要在理解诊断结果后才能采取行动。我们提出ExplainableCropAI，一个用于水稻病害诊断的决策支持系统，它在ConvNeXt V2-Tiny分类器上附加了三条互补的解释通道：Grad-CAM++视觉归因、经温度缩放校准的蒙特卡洛Dropout不确定性，以及由Phi-3-mini在接地约束下生成的自然语言解释——该约束将模型限制为从预测结果、归因图和精选农学知识库中提取的类型化字段。各组件均非新创；贡献在于将其集成到单一流水线中，并测量该集成旨在提供的特性。在来自Paddy Doctor数据集（10,407张图像，九种病害及健康）的1,540张图像上，经泄漏控制的留出测试划分显示，分类器达到97.5%的准确率（95% CI 96.7–98.3）和96.9%的宏F1值。在相同设置下，使用基于模型的忠实性指标比较了六种归因方法；校准后的置信度以0.93的AUROC区分正确与错误预测，将置信度最低的10%案例转交专家可消除全部错误的74%。在五种严重程度的八种图像损坏下，模型保持89.7%的平均准确率，而相同训练的六个基线为72–78%，且其转交率随退化程度上升。在来自孟加拉国和印度的两个独立采集的水稻病害数据集上进行零样本评估显示，跨数据集偏移下准确率大约减半，置信度不再能识别错误，而每类五十张本地图像可将准确率恢复至90%以上。代码、划分文件、训练权重和评估脚本均已发布。","International Journal of Computational Intelligence Systems",84,[190],{"name":187,"url":184},[20,192,99,193,194],"农业人工智能","可解释AI","作物病害诊断",{"code":22,"slug":23,"name":24,"path":25,"parent_code":26},[197],{"code":104,"slug":105,"name":99,"path":106,"parent_code":107},[199,200],"ExplainableCropAI 水稻病害","Paddy Doctor 数据集","ExplainableCropAI水稻病害-5613","10.1007\u002Fs44196-026-01654-9",{"doi":202,"openalex_id":204,"authors":205,"venue":187,"cited_by_count":16,"oa_url":184,"card":215,"direction":50,"ingested_from":51},"W7166554515",[206,208,209,212,214],{"name":207,"orcid":12},"Santosh Kumar",{"name":207,"orcid":12},{"name":210,"orcid":211},"Durgesh Nandan","https:\u002F\u002Forcid.org\u002F0000-0002-9762-3559",{"name":213,"orcid":12},"Sumit Kumar",{"name":213,"orcid":12},{"tldr":216,"method":217,"finding":218,"direction":50,"opportunity":219},"构建可解释水稻病害诊断系统，融合归因、不确定性与自然语言解释。","ConvNeXt V2-Tiny分类器，Grad-CAM++、MC Dropou","测试准确率97.5%，低置信转诊消除74%错误，跨数据集精度减半。","跨数据集置信失效，少量本地样本微调可恢复精度，值得研究田间自适应与主动学习。","2026-10-22T23:31:14.582261Z",{"id":222,"title":223,"url":224,"summary":225,"summary_zh":226,"source_name":227,"source_url":224,"published_at":228,"category":11,"cover_url":12,"hotness":13,"is_selected":14,"embedding_ready":15,"score":229,"sources":230,"tags":232,"primary_topic":236,"secondary_topics":237,"search_phrases":238,"slug":241,"view_count":16,"doi":242,"paper":243,"created_at":259},5611,"RGB UAV Imagery for Crop Monitoring and Decision Support in Precision Agriculture","https:\u002F\u002Fdoi.org\u002F10.3390\u002Fagriengineering8100448","Digitalization and data-driven decision-making are transforming modern agriculture, where rapid and high-resolution assessment of crop condition is essential for optimizing field interventions. Multispectral approaches such as satellite-derived normalized difference vegetation index are widely used; however, their spatial resolution and temporal availability may limit the detection of fine-scale variability within agricultural fields. This study investigates the capability of unmanned aerial vehicle (UAV)-acquired red, green, and blue (RGB) imagery to provide relevant information regarding crop condition. Two autumn crops, wheat and rapeseed, located in Olt County, Romania, were analyzed. The acquired images were processed photogrammetrically to generate orthomosaics, followed by the extraction of vegetation indices derived from the visible spectrum, namely Excess Green Index and Visible Atmospherically Resistant Index. Both indices provided meaningful information on vegetation condition, with the latter exhibiting enhanced sensitivity to intra-field variability. The UAV-derived Excess Green Index and Visible Atmospherically Resistant Index maps revealed spatial variability that was not equally distinguishable in the satellite Normalized Difference Vegetation Index data, particularly for the wheat crop. Individually, the two RGB vegetation indices classified approximately 16% of each analyzed field as vulnerable. These vulnerable regions represent relative, image-based spectral anomalies rather than independently validated indicators of physiological crop stress, nutrient deficiency, or yield loss. Following rule-based fusion and the removal of isolated detections, 21.55% of the wheat field and 19.75% of the rapeseed field were identified as spatially coherent vulnerable areas. The resulting maps were subsequently converted into machinery-oriented candidate intervention zones, supporting an illustrative simulation of how spatially differentiated field information could be integrated into conventional agricultural equipment. These findings demonstrate that UAV-based RGB imagery can provide spatially explicit information for crop monitoring and can support the identification of candidate areas for subsequent field verification and precision-agriculture decision-making.","数字化与数据驱动的决策正在改变现代农业，其中对作物状况进行快速、高分辨率评估对于优化田间干预措施至关重要。卫星衍生的归一化差异植被指数（NDVI）等多光谱方法已被广泛使用；然而，其空间分辨率和时间可获得性可能限制了对农田内部细尺度变异的检测。本研究探讨了无人机（UAV）获取的红、绿、蓝（RGB）影像在提供作物状况相关信息方面的能力。研究分析了位于罗马尼亚奥尔特县的两种秋季作物——小麦和油菜。对获取的影像进行摄影测量处理以生成正射镶嵌图，随后提取基于可见光谱的植被指数，即超绿指数（Excess Green Index）和可见光大气阻抗指数（Visible Atmospherically Resistant Index）。两种指数均提供了有关植被状况的有意义信息，其中后者对田内变异表现出更高的敏感性。无人机衍生的超绿指数和可见光大气阻抗指数图揭示了卫星归一化差异植被指数数据中未能同等区分出的空间变异，尤其对小麦作物而言。单独来看，这两种RGB植被指数将每个分析田块约16%的区域归类为脆弱区。这些脆弱区域代表的是基于影像的相对光谱异常，而非经过独立验证的作物生理胁迫、养分缺乏或产量损失指标。经过基于规则的融合并去除孤立检测后，小麦田的21.55%和油菜田的19.75%被识别为空间连贯的脆弱区域。随后将所得地图转换为面向机械的候选干预区，支持了一项示意性模拟，展示如何将空间差异化的田间信息整合到常规农业设备中。这些发现表明，基于无人机的RGB影像可为作物监测提供空间明确的信息，并能够支持识别供后续田间验证和精准农业决策的候选区域。","AgriEngineering","2026-10-21T00:00:00Z",68,[231],{"name":227,"url":224},[20,233,74,234,235],"无人机","遥感","作物监测",{"code":22,"slug":23,"name":24,"path":25,"parent_code":26},[],[239,240],"RGB无人机 作物监测","Olt 小麦 油菜","RGB无人机作物监测-5611","10.3390\u002Fagriengineering8100448",{"doi":242,"openalex_id":244,"authors":245,"venue":227,"cited_by_count":16,"oa_url":224,"card":252,"direction":258,"ingested_from":51},"W7223174201",[246,249],{"name":247,"orcid":248},"Robert Ciobanu","https:\u002F\u002Forcid.org\u002F0000-0001-7518-974X",{"name":250,"orcid":251},"Emilia Georgiana Prisăcariu","https:\u002F\u002Forcid.org\u002F0009-0005-2437-7288",{"tldr":253,"method":254,"finding":255,"direction":256,"opportunity":257},"用无人机RGB影像提取植被指数监测小麦和油菜长势并生成精准干预区。","无人机RGB影像摄影测量生成正射图，计算ExG与VARI指数并规则融合。","RGB指数可识别卫星NDVI难分辨的田间细微变异，约20%田块为脆弱区。","农业遥感与作物表型","可结合地面验证与多源遥感，将RGB脆弱区图转化为可验证的胁迫诊断与变量作业决策。","数字乡村与农业信息化","2026-10-22T23:30:52.745269Z",{"id":261,"title":262,"url":263,"summary":264,"summary_zh":265,"source_name":266,"source_url":263,"published_at":228,"category":11,"cover_url":12,"hotness":13,"is_selected":14,"embedding_ready":15,"score":267,"sources":268,"tags":270,"primary_topic":272,"secondary_topics":273,"search_phrases":278,"slug":281,"view_count":16,"doi":282,"paper":283,"created_at":312},5603,"AI-Enabled Smart and Precision Agriculture: Smart Sensors, IoT, Deep Learning, Computer Vision, and Decision Support: A Comprehensive Survey","https:\u002F\u002Fdoi.org\u002F10.20944\u002Fpreprints202610.2112.v1","Artificial intelligence (AI) is increasingly integrated with smart sensors, Internet of Things (IoT) networks, machine learning, deep learning, and computer vision to support site-specific agricultural management. However, strong performance on computational benchmarks does not necessarily translate into reliable decisions, effective field interventions, or measurable improvements in agricultural productivity and resource efficiency. This comprehensive narrative survey critically synthesizes advances in agricultural sensing, IoT architectures, data processing, multimodal sensor fusion, predictive modeling, computer vision, decision optimization, and automated actuation. Particular emphasis is placed on the relationships between measurement reliability, model generalization, decision feasibility, intervention accuracy, and verified agricultural outcomes. The reviewed literature demonstrates promising capabilities in crop monitoring, yield prediction, disease and weed recognition, precision irrigation, and selected automated operations. Nevertheless, heterogeneous evaluation protocols, limited independent field validation, sensor calibration challenges, distribution shifts, interoperability barriers, and deployment costs constrain operational reliability and generalizability. To address these interconnected challenges, the survey proposes a literature-informed reference architecture linking agricultural observations, data integration, AI inference, constrained decision-making, physical interventions, and outcome feedback, supported by uncertainty assessment, human oversight, and cross-layer assurance. The proposed architecture is conceptual and has not been empirically validated. Future research priorities include reproducible field evaluations, representative agricultural datasets, uncertainty-aware and explainable AI, resource-efficient computing, interoperable infrastructures, and economically inclusive deployment. The review concludes that the practical value of AI-enabled precision agriculture should be evaluated through reliable interventions, sustainable resource utilization, and demonstrated agricultural outcomes rather than predictive performance alone.","人工智能（AI）正日益与智能传感器、物联网（IoT）网络、机器学习、深度学习及计算机视觉相融合，以支撑因地制宜的精准农业管理。然而，在计算基准上的优异表现并不必然转化为可靠的决策、有效的田间干预，或农业生产率与资源效率方面可衡量的提升。本综述性研究系统梳理并批判性综合了农业传感、物联网架构、数据处理、多模态传感器融合、预测建模、计算机视觉、决策优化及自动执行等方面的进展，重点探讨测量可靠性、模型泛化能力、决策可行性、干预精度与经核实的农业成效之间的关系。所综述文献表明，上述技术在作物监测、产量预测、病害与杂草识别、精准灌溉及部分自动化作业中展现出良好前景。然而，评估方案异质、独立田间验证有限、传感器校准难题、分布偏移、互操作性障碍及部署成本等因素制约了运行可靠性与泛化能力。为应对这些相互关联的挑战，本综述提出一种基于文献的参考架构，将农业观测、数据集成、AI推理、受约束决策、物理干预与成效反馈相衔接，并以不确定性评估、人工监督与跨层保障为支撑。该架构为概念性框架，尚未经实证验证。未来研究重点包括可复现的田间评估、具有代表性的农业数据集、不确定性感知与可解释AI、资源高效计算、可互操作基础设施及经济包容性部署。本综述认为，AI赋能精准农业的实际价值应通过可靠的干预措施、可持续的资源利用及经证实的农业成效来评估，而非仅凭预测性能。","Preprints.org",70,[269],{"name":266,"url":263},[20,192,271,82,74],"计算机视觉",{"code":22,"slug":23,"name":24,"path":25,"parent_code":26},[274],{"code":78,"slug":275,"name":276,"path":277,"parent_code":26},"agricultural-iot","物联网传感与边缘计算","\u002Ftopic\u002Fdigital-agriculture\u002Fagricultural-iot",[279,280],"农业人工智能 计算机视觉 智慧农业 精准农业","农业人工智能 计算机视觉","农业人工智能计算机视觉智慧农业精准农业-5603","10.20944\u002Fpreprints202610.2112.v1",{"doi":282,"openalex_id":284,"authors":285,"venue":266,"cited_by_count":16,"oa_url":263,"card":306,"direction":310,"ingested_from":51},"W7223175252",[286,289,292,295,297,299,302,304],{"name":287,"orcid":288},"IBRAHIM ADABARA","https:\u002F\u002Forcid.org\u002F0000-0001-8949-7540",{"name":290,"orcid":291},"Bashir Olaniyi Sadiq","https:\u002F\u002Forcid.org\u002F0000-0001-7868-4076",{"name":293,"orcid":294},"Aliyu Nuhu Shuaibu","https:\u002F\u002Forcid.org\u002F0000-0002-5171-675X",{"name":296,"orcid":12},"Yale Ibrahim Danjuma",{"name":298,"orcid":12},"Ainamukama Anold",{"name":300,"orcid":301},"Mutebi Joe","https:\u002F\u002Forcid.org\u002F0000-0002-2757-3875",{"name":303,"orcid":12},"Elhadi Abdalla A. Arbab",{"name":305,"orcid":12},"Brian Mugisha",{"tldr":307,"method":308,"finding":309,"direction":310,"opportunity":311},"综述AI赋能精准农业的传感、物联网、深度学习与决策支持，并提出参考架构。","叙述性综述，整合传感器、IoT、多模态融合、计算机视觉与决策优化文献。","基准性能不等于田间可靠决策，评估协议异质、独立验证不足制约落地。","智慧农业 \u002F 农业物联网","可研究不确定性感知与可解释AI、可复现田间评估及跨层保障架构的实证验证。","2026-10-22T23:30:19.310160Z",{"id":314,"title":315,"url":316,"summary":317,"summary_zh":318,"source_name":319,"source_url":316,"published_at":228,"category":11,"cover_url":12,"hotness":13,"is_selected":14,"embedding_ready":15,"score":320,"sources":321,"tags":323,"primary_topic":327,"secondary_topics":333,"search_phrases":335,"slug":338,"view_count":16,"doi":339,"paper":340,"created_at":354},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)",69,[322],{"name":319,"url":316},[20,192,324,325,326],"小农户","植物病害识别","可解释性",{"code":328,"slug":329,"name":330,"path":331,"parent_code":332},"G2","farm-services","社会化服务与规模经营","\u002Ftopic\u002Frural-services\u002Ffarm-services","G",[334],{"code":22,"slug":23,"name":24,"path":25,"parent_code":26},[336,337],"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":339,"openalex_id":341,"authors":342,"venue":319,"cited_by_count":16,"oa_url":316,"card":349,"direction":50,"ingested_from":51},"W7223137323",[343,345,347],{"name":344,"orcid":12},"Akshay Ramgoolam",{"name":346,"orcid":12},"Priya Bundhoo Appadoo",{"name":348,"orcid":12},"Jean-Noël LEspérance",{"tldr":350,"method":351,"finding":352,"direction":50,"opportunity":353},"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。","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":356,"title":357,"url":358,"summary":359,"summary_zh":360,"source_name":319,"source_url":358,"published_at":228,"category":11,"cover_url":12,"hotness":13,"is_selected":14,"embedding_ready":15,"score":320,"sources":361,"tags":363,"primary_topic":367,"secondary_topics":372,"search_phrases":374,"slug":377,"view_count":16,"doi":378,"paper":379,"created_at":391},5591,"Hierarchical Conformal Risk Control for Artificial Intelligence Agriculture Irrigation Scheduling Under Unmodeled Infiltration: A Computer Safety-Guarantee Evaluation","https:\u002F\u002Fdoi.org\u002F10.5281\u002Fzenodo.23543665","Irrigation scheduling algorithms that rely on soil-moisture forecasts are vulnerable to unmodeled infiltration, the process by which water bypasses the root zone through preferential pathways that resistance-based soil models cannot represent. The framework is evaluated through a replication-based design study that characterises the operating characteristics of the before-after-control-impact estimator used to audit scheduler performance, under a declared variance structure with between-cluster standard deviation 6.0, residual standard deviation 4.5 and intraclass correlation 0.64. Power was estimated across 2000 design cells spanning cluster counts of 8, 16, 24 and 32 and effect sizes of 3, 5 and 8 percentage points. The analysis indicates that 54 clusters are required to detect a 5-percentage-point effect at 80% power, and that the minimum detectable effect declines steeply as cluster count rises from 8 to 32. The contribution is a formal safety guarantee for irrigation decision support that remains valid when the infiltration model is wrong, together with an audit design that makes the guarantee empirically checkable.","依赖土壤湿度预报的灌溉调度算法容易受到未建模入渗的影响，即水分通过优先路径绕过根区的过程，而基于阻力的土壤模型无法表征这一过程。该框架通过一项基于重复的设计研究进行评估，该研究刻画了用于审计调度器性能的前后对照影响估计量的运行特性，其所依据的声明方差结构为：组间标准差6.0、残差标准差4.5、组内相关系数0.64。功效估计覆盖2000个设计单元，涵盖聚类数8、16、24和32以及效应量3、5和8个百分点。分析表明，在80%功效下检测5个百分点效应需要54个聚类，且最小可检测效应随聚类数从8增至32而急剧下降。本文的贡献在于为灌溉决策支持提供了一种正式的安全保证，该保证在入渗模型错误时仍然有效，同时提出了一种使该保证可经实证检验的审计设计。",[362],{"name":319,"url":358},[20,192,364,365,366],"灌溉决策","安全保证","试验设计",{"code":368,"slug":369,"name":370,"path":371,"parent_code":114},"H3","irrigation","水资源与灌溉","\u002Ftopic\u002Fenvironment\u002Firrigation",[373],{"code":22,"slug":23,"name":24,"path":25,"parent_code":26},[375,376],"灌溉调度 安全保证","农业人工智能 安全保证 智慧农业 灌溉决策","灌溉调度安全保证-5591","10.5281\u002Fzenodo.23543665",{"doi":378,"openalex_id":380,"authors":381,"venue":319,"cited_by_count":16,"oa_url":358,"card":386,"direction":50,"ingested_from":51},"W7223131205",[382,384],{"name":383,"orcid":12},"Nassima Benkhelifa",{"name":385,"orcid":12},"Yacine Aït Ahmed",{"tldr":387,"method":388,"finding":389,"direction":50,"opportunity":390},"提出分层保形风险控制，为未建模入渗下的灌溉调度提供安全保证。","分层保形风险控制与基于重复设计的BACI审计，2000个设计单元功效分析。","检测5个百分点效应需54个集群，最小可检测效应随集群数增加陡降。","可将保形风险控制扩展到其他未建模扰动下的农业决策安全保证与审计设计。","2026-10-22T11:21:10.283561Z"]