[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"item-3229":3,"related-3229":36},{"id":4,"title":5,"url":6,"summary":7,"summary_zh":8,"content":8,"source_name":9,"source_url":8,"published_at":10,"category":11,"cover_url":8,"hotness":12,"is_selected":13,"score":14,"score_detail":15,"sources":22,"tags":24,"search_phrases":30,"slug":33,"view_count":34,"doi":8,"paper":8,"created_at":35},3229,"乔府大院第十一届开镰节系列活动启幕——黑龙江省农科院为乔府大院授牌\"科技包联示范基地\"","https:\u002F\u002Fwww.hlj.news.cn\u002F20260922\u002F30b3409efc004627ab756efa3334249e\u002Fc.html","9-22 新华网黑龙江频道报道：9-22 由黑龙江省农科院、哈尔滨市农业农村局、五常市农业农村局、亿林学堂联合主办，乔府大院承办的黑龙江省农科院科技包联哈尔滨市现场观摩会、首届亿林学堂世界商业武道大赛、第十一届乔府大院开镰节系列活动，在五常市乔府大院现代农业产业园启幕。依托黑龙江省\"科技包联、赋能增粮\"工作部署，黑龙江省农业科学院专家团队长期扎根乔府大院田间一线，围绕精准施肥、富硒稻米种植、良种选育研发、智慧农业应用四大领域持续技术赋能。活动现场，黑龙江省农业科学院正式授予乔府大院\"科技包联示范基地\"称号。最新数据显示，截至2026年上半年，黑龙江省定制农业主体数量发展到445家，种植定制基地面积达到179.9万亩，实现签约销售额55.5亿元，同比增长85%。",null,"新华网黑龙江频道","2026-09-22T00:00:00Z","报道",10,false,74,{"impact":16,"substance":17,"depth":18,"authority":19,"freshness":12,"relevant":20,"comment":21},18,20,14,12,1,"省级科技包联授牌叠加定制农业最新数据，信息增量足、时效性强，但属企业承办活动，影响层级限于省级与细分行业。",[23],{"name":9,"url":6},[25,26,27,28,29],"智慧农业","科技包联","五常大米","定制农业","富硒稻米",[31,32],"黑龙江省农科院 乔府大院 科技包联示范基地","五常大米 开镰节 定制农业","黑龙江省农科院乔府大院科技包联示范基地-3229",0,"2026-09-23T00:04:30.450309Z",{"total":37,"page":20,"page_size":37,"items":38},6,[39,74,105,132,160,187],{"id":40,"title":41,"url":42,"summary":43,"summary_zh":8,"content":8,"source_name":44,"source_url":8,"published_at":45,"category":46,"cover_url":8,"hotness":12,"is_selected":13,"score":47,"score_detail":48,"sources":53,"tags":55,"search_phrases":60,"slug":63,"view_count":34,"doi":8,"paper":64,"created_at":73},3253,"Visuomotor Robotic Pruning in Planar Orchills Using Hybrid Reinforcement Learning（基于混合强化学习的平面果园视觉运动机器人修剪）","https:\u002F\u002Farxivtldr.org\u002Fabs\u002F2609.24906","arXiv发表研究：休眠期树木修剪是劳动密集型但对维持现代高产果园至关重要的工作。本文针对现代平面树形训练系统（V型棚架苹果、UFO樱桃）的修剪任务，提出端到端管线学习闭环视觉运动控制器。控制器完全使用仿真和合成生成数据进行训练，以零样本方式部署到真实果园。管线包括平面果园树网格的合成生成、基于物理的果园仿真器构建、通过运动规划自动收集成功修剪轨迹，以及结合离线演示与在线仿真推出相结合的新型混合强化学习算法。控制器使用腕部相机的光流输入，避免完整三维重建需求，连续引导切割器穿越杂乱分支环境到指定切割点并在真实果园38次物理试验中展示零样本仿真到真实迁移。在V-Trellis苹果上达到49.9%成功率，UFO樱桃上46.0%。","arXiv (Abhinav Jain, Cindy Grimm, Stefan Lee)","2026-09-20T00:00:00Z","论文",80,{"impact":16,"substance":49,"depth":16,"authority":50,"freshness":51,"relevant":20,"comment":52},22,13,9,"仿真到真实零样本迁移的果园修剪机器人研究，方法新颖、数据扎实，对智慧果园机械化有参考价值。",[54],{"name":44,"url":42},[25,56,57,58,59],"农业人工智能","农业机器人","强化学习","果园管理",[61,62],"V-Trellis 苹果 机器人修剪","UFO 樱桃 零样本迁移","V-Trellis苹果机器人修剪-3253",{"doi":8,"openalex_id":8,"authors":65,"venue":8,"cited_by_count":34,"oa_url":8,"card":66,"direction":70,"ingested_from":72},[],{"tldr":67,"method":68,"finding":69,"direction":70,"opportunity":71},"提出端到端视觉运动控制器，用仿真合成数据训练，零样本迁移到真实果园完成修剪。","合成数据生成、物理仿真器、运动规划收集轨迹、混合强化学习、腕部相机光流。","真实果园38次试验零样本迁移成功，苹果49.9%、樱桃46.0%成功率。","农业人工智能与决策模型","可探索多季节、多树种泛化及真实数据微调，提升复杂冠层下的鲁棒性。","agent","2026-09-23T00:04:33.854148Z",{"id":75,"title":76,"url":77,"summary":78,"summary_zh":8,"content":8,"source_name":79,"source_url":8,"published_at":10,"category":46,"cover_url":8,"hotness":12,"is_selected":80,"score":81,"score_detail":82,"sources":88,"tags":90,"search_phrases":93,"slug":96,"view_count":34,"doi":8,"paper":97,"created_at":104},3252,"具身智能农业机器人关键技术与发展趋势——上海大学苗中华等\"感知-决策-模拟-进化-诊断\"五位一体研究框架","http:\u002F\u002Fwww.qikanzj.com\u002Fhek\u002Fnyjxxb\u002Fmulu\u002F455891.html","智慧农业(中英文)期刊发表上海大学苗中华、朱子煜、张伟、薛振锋、孙腾、张异凡、谢涛、何创新、李楠，山东农业大学苑进，北京市农林科学院赵春江，上海交通大学刘成良团队论文：全球农业智能化转型背景下，传统农业机器人面临非结构化环境、作业对象不确定性及控制参数时变等技术瓶颈，导致农业机器人实用效果差，产业化落地困难。具身智能作为实现通用人工智能的关键路径，为突破上述瓶颈提供了新范式。本文系统阐述具身智能理论驱动的农业机器人技术体系与发展路径，创新性构建\"感知-决策-模拟-进化-诊断\"五位一体研究框架。具身感知聚焦开放环境场景理解、作业目标主动感知及多模态融合感知；具身决策涵盖端到端导航、大模型驱动的机械臂操作及多机协同算法；具身模拟着力于高保真场景重建、生成及虚实迁移技术；具身进化整合无监督学习、强化学习及模仿学习机制；具身诊断构建作业状态监测与质量评估体系。","智慧农业(中英文)·上海大学",true,91,{"impact":83,"substance":84,"depth":85,"authority":86,"freshness":12,"relevant":20,"comment":87},24,23,19,15,"核心期刊论文，提出具身智能驱动的农业机器人五位一体框架，方法体系新颖、团队权威，对智慧农业技术路线有较强参考价值。",[89],{"name":79,"url":77},[25,56,57,91,92],"具身智能","多模态感知",[94,95],"上海大学 苗中华 农业机器人","具身智能 农业机器人 五位一体","上海大学苗中华农业机器人-3252",{"doi":8,"openalex_id":8,"authors":98,"venue":8,"cited_by_count":34,"oa_url":8,"card":99,"direction":70,"ingested_from":72},[],{"tldr":100,"method":101,"finding":102,"direction":70,"opportunity":103},"系统阐述具身智能驱动的农业机器人技术体系，提出五位一体研究框架。","构建'感知-决策-模拟-进化-诊断'五位一体框架，综述关键技术。","具身智能可突破非结构化环境等瓶颈，为农业机器人提供新范式。","可探索具身智能在农业非结构化场景的落地验证与虚实迁移效率提升。","2026-09-23T00:04:33.744630Z",{"id":106,"title":107,"url":108,"summary":109,"summary_zh":8,"content":8,"source_name":110,"source_url":8,"published_at":10,"category":46,"cover_url":8,"hotness":12,"is_selected":13,"score":111,"score_detail":112,"sources":114,"tags":116,"search_phrases":119,"slug":122,"view_count":34,"doi":8,"paper":123,"created_at":131},3251,"农机化研究2026(8)：基于MobileNetV4-DSFPN的芍药田间机器人视觉导航——安徽理工+皖西学院+中科院合肥物质科学研究院 mIoU 94.11% FPS 23.1","http:\u002F\u002Fwww.qikanvip.com\u002Fqkml\u002F166165.html","农机化研究2026年第8期发表安徽理工大学徐善永、程军辉、张俊卿，皖西学院邢雪景，中国科学院合肥物质科学研究院联合论文：精准分割田间可行驶区域并实时提取导航线是实现农业机器人在田间自主作业的关键环节。针对芍药田间背景复杂、现有语义模型计算复杂度高、实时性差等问题，提出轻量化的MobileNetV4-DSFPN语义分割模型，采用改进MobileNetV4作为高效编码器；解码器部分构建基于深度可分离卷积的轻量级特征金字塔网络（DSFPN）。在芍药田间路径数据集上类别平均像素准确率（mPA）和平均交并比（mIoU）分别达到97.16%、94.11%；导航线的平均横向偏差为0.231像素，平均角度偏差为0.842°；车载计算机上平均推理速度达到23.1 FPS。","农机化研究",78,{"impact":86,"substance":49,"depth":16,"authority":50,"freshness":12,"relevant":20,"comment":113},"核心期刊论文，提出轻量化语义分割模型并给出mIoU 94.11%、23.1 FPS等实测数据，方法新颖、结论可靠，对农业机器人视觉导航有实质参考价值。",[115],{"name":110,"url":108},[25,56,57,117,118],"芍药","视觉导航",[120,121],"安徽理工大学 芍药 田间机器人","MobileNetV4 DSFPN 语义分割","安徽理工大学芍药田间机器人-3251",{"doi":8,"openalex_id":8,"authors":124,"venue":8,"cited_by_count":34,"oa_url":8,"card":125,"direction":129,"ingested_from":72},[],{"tldr":126,"method":127,"finding":128,"direction":129,"opportunity":130},"提出轻量MobileNetV4-DSFPN分割芍药田间可行驶区域并提取导航线。","改进MobileNetV4编码器+深度可分离卷积特征金字塔，芍药田间路径数据集。","mIoU达94.11%，横向偏差0.231像素，车载推理23.1 FPS。","智慧农业 \u002F 农业物联网","可探索轻量分割模型在复杂田间多作物、多光照下的泛化与边缘部署优化。","2026-09-23T00:04:33.588064Z",{"id":133,"title":134,"url":135,"summary":136,"summary_zh":8,"content":8,"source_name":110,"source_url":8,"published_at":10,"category":46,"cover_url":8,"hotness":12,"is_selected":13,"score":137,"score_detail":138,"sources":141,"tags":143,"search_phrases":147,"slug":150,"view_count":34,"doi":8,"paper":151,"created_at":159},3250,"农机化研究2026(11)：基于多源感知与语义分割的果树行识别与路径跟踪控制——郑州工业应用技术学院许洋洋等 DeepLabv3+ F1达0.94","http:\u002F\u002Fwww.qikanvip.com\u002Fqkml\u002F166268.html","农机化研究2026年第11期发表郑州工业应用技术学院许洋洋、柴亚珂、王莹、牛瑞利团队论文：针对果园中裸露土壤、遮蔽物干扰和冠层阴影等因素导致导航路径难以稳定提取的问题，提出融合多源感知与语义分割的自走式植保机果树识别与路径控制方法。首先利用无人机获取多光谱影像并生成正射影像（DOM）与数字表面模型（DSM），计算归一化差异绿度指数（NDGI），增强果树冠层与背景的可分性；然后采用U-Net\u002FDeepLabv3+语义分割网络在保留空间上下文信息的同时实现果树行\u002F背景端到端分割。田间对比试验结果表明，DeepLabv3+整体分割F1可达0.94，每样区生成导航线断裂1.2次，且转向工况下横向误差RMSE为0.11 m，显著优于传统阈值分割和SVM方法。",79,{"impact":139,"substance":49,"depth":16,"authority":50,"freshness":12,"relevant":20,"comment":140},16,"多源遥感与语义分割结合的果园导航研究，F1达0.94且横向误差0.11米，方法新颖、数据扎实，具备行业参考价值。",[142],{"name":110,"url":135},[25,56,144,145,146],"农机导航","遥感","果园植保",[148,149],"郑州工业应用技术学院 果树行识别","DeepLabv3 果园导航 路径跟踪","郑州工业应用技术学院果树行识别-3250",{"doi":8,"openalex_id":8,"authors":152,"venue":8,"cited_by_count":34,"oa_url":8,"card":153,"direction":157,"ingested_from":72},[],{"tldr":154,"method":155,"finding":156,"direction":157,"opportunity":158},"提出融合无人机多光谱与语义分割的果树行识别与路径跟踪方法，实现植保机自主导航。","无人机多光谱DOM\u002FDSM与NDGI，U-Net\u002FDeepLabv3+语义分割，","DeepLabv3+分割F1达0.94，导航线断裂1.2次，转向横向误差RMSE 0.11 m，优于","农业遥感与作物表型","可探索多光谱与DSM特征融合的轻量化分割模型，并迁移至多树种、多季节果园导航。","2026-09-23T00:04:33.496233Z",{"id":161,"title":162,"url":163,"summary":164,"summary_zh":8,"content":8,"source_name":165,"source_url":8,"published_at":10,"category":46,"cover_url":8,"hotness":12,"is_selected":13,"score":166,"score_detail":167,"sources":169,"tags":171,"search_phrases":175,"slug":178,"view_count":34,"doi":8,"paper":179,"created_at":186},3249,"设施番茄采摘机器人识别定位与采摘方法研究——江苏大学雷志龙等 基于改进YOLO v5-HSV融合算法识别准确率95.01%","http:\u002F\u002Fwww.qikanzj.com\u002Fhek\u002Fnyjxxb\u002Fmulu\u002F455758.html","智慧农业(中英文)期刊发表江苏大学雷志龙、刘畅、王权团队研究：针对设施单果番茄采摘需求，设计一款设施番茄智能采摘平台，主要由升降机构、采摘机构、识别与定位系统等部分组成。平台整体结构由低压一体化伺服滚珠丝杠副升降机构、六轴协作机械臂和力控末端执行器组成。基于改进YOLO v5-HSV融合算法来识别检测，通过对H分量进行图像阈值分割，提高对成熟目标果实识别的准确率，有效排除未成熟番茄和枝叶背景的干扰；通过眼在手外的标定方法，使用ZED双目相机进行定位。搭建的设施番茄采摘机样机平台在现场采摘试验中识别准确率达到95.01%，采摘成功率为87.96%，单果平均采摘时间为14.56s。","智慧农业(中英文)·江苏大学",82,{"impact":16,"substance":49,"depth":16,"authority":18,"freshness":12,"relevant":20,"comment":168},"核心期刊论文，方法有改进、数据完整，识别准确率与采摘成功率等指标明确，对设施农业智能装备研发有参考价值。",[170],{"name":165,"url":163},[25,56,172,173,174],"采摘机器人","机器视觉","设施番茄",[176,177],"江苏大学 设施番茄 采摘机器人","YOLO v5 番茄 识别定位","江苏大学设施番茄采摘机器人-3249",{"doi":8,"openalex_id":8,"authors":180,"venue":8,"cited_by_count":34,"oa_url":8,"card":181,"direction":70,"ingested_from":72},[],{"tldr":182,"method":183,"finding":184,"direction":70,"opportunity":185},"设计设施番茄智能采摘平台，融合改进YOLO v5与HSV实现识别定位与采摘。","改进YOLO v5-HSV融合算法、ZED双目相机眼在手外标定、六轴机械臂力控末","识别准确率95.01%，采摘成功率87.96%，单果平均采摘时间14.56秒。","可探索多果簇、遮挡与弱光环境下识别定位鲁棒性，并优化采摘效率与末端力控。","2026-09-23T00:04:33.401768Z",{"id":188,"title":189,"url":190,"summary":191,"summary_zh":8,"content":8,"source_name":192,"source_url":8,"published_at":10,"category":46,"cover_url":8,"hotness":12,"is_selected":13,"score":193,"score_detail":194,"sources":196,"tags":198,"search_phrases":202,"slug":205,"view_count":34,"doi":8,"paper":206,"created_at":213},3248,"Crop recommendation in precision agriculture: a systematic literature review of methods, trends, and challenges（精准农业中的作物推荐：方法、趋势与挑战系统综述）","https:\u002F\u002Fm2.mtmt.hu\u002Fapi\u002Fpublication\u002F37471110","MDPI 发表精准农业作物推荐方法系统综述：从183项研究中严格筛选129篇发表于2020-2026年的文章，使用PRISMA协议进行综合分析。研究表明集成学习方法（特别是随机森林和XGBoost）在各种农业数据集的预测性能上具有强大能力；支持向量机、决策树、k近邻等传统ML方法仍被广泛使用；同时CNN和LSTM被用于遥感和时间相关农业分析。最常用的数据集来源是Kaggle，典型输入包括土壤养分（NPK）、土壤pH、天气条件和NDVI、EVI等卫星指数。研究主要研究空白：有限的实时部署、低多数据源集成、低跨区域验证、低模型可解释性。研究表明可扩展、可解释的AI系统对农业实际应用具有重要意义。","MDPI",81,{"impact":16,"substance":49,"depth":16,"authority":50,"freshness":12,"relevant":20,"comment":195},"基于PRISMA的129篇文献系统综述，梳理作物推荐主流方法与四大研究空白，对农业AI落地有参考价值。",[197],{"name":192,"url":190},[25,56,199,200,201],"机器学习","作物推荐","遥感监测",[203,204],"精准农业 作物推荐 系统综述","XGBoost 随机森林 作物推荐","精准农业作物推荐系统综述-3248",{"doi":8,"openalex_id":8,"authors":207,"venue":8,"cited_by_count":34,"oa_url":8,"card":208,"direction":70,"ingested_from":72},[],{"tldr":209,"method":210,"finding":211,"direction":70,"opportunity":212},"系统综述129篇2020-2026年文献，梳理精准农业作物推荐的方法、趋势与挑战。","PRISMA协议系统综述，分析183项研究筛选出的129篇文献。","集成学习（随机森林、XGBoost）表现最强，主要空白为实时部署、多源集成、跨区域验证与可解释性。","可探索可解释、可跨区域泛化的实时作物推荐系统，并融合多源遥感与物联网数据。","2026-09-23T00:04:33.331160Z"]