[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"item-2277":3,"related-2277":86},{"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":9,"secondary_topics":32,"classification":33,"search_phrases":61,"slug":64,"view_count":65,"doi":66,"paper":67,"created_at":85},2277,"An improved RT-DETR approach for real-time detection of densely occluded blueberries and micro-calyxes","https:\u002F\u002Fdoi.org\u002F10.1016\u002Fj.compag.2026.112424","Accurate detection of blueberries and their micro-scale calyx regions is important for visual perception in robotic blueberry harvesting, especially in densely clustered scenes where fruits are frequently occluded and adjacent to each other. However, blueberry calyx detection remains challenging because the front calyx occupies only a small image region and is easily affected by background interference, illumination variation, and feature attenuation during deep feature extraction. To address these challenges, this study proposes MFE-DETR, an improved RT-DETR-based real-time detection model for densely occluded blueberries and micro-calyxes. A Micro-Feature Enhanced Attention (MFEA) module is introduced to strengthen the representation of weak calyx features by combining heterogeneous local receptive fields with a spatial prior bias. The proposed module is designed to enhance the association between micro-calyx regions and the surrounding fruit context while maintaining the end-to-end detection framework of RT-DETR. Experimental results on a self-constructed blueberry image dataset show that MFE-DETR achieved an overall mAP@0.5 of 94.2 %, with a Front calyx AP@0.5 of 91.5 %. In addition, deployment experiments on an NVIDIA Jetson AGX Orin platform showed that the proposed model achieved real-time edge-side inference under the tested configuration, suggesting its potential as a visual perception module for subsequent robotic blueberry harvesting systems.","准确检测蓝莓及其微尺度花萼区域对于机器人蓝莓采摘中的视觉感知具有重要意义，尤其是在果实频繁被遮挡且相互邻近的密集簇生场景中。然而，蓝莓花萼检测仍面临挑战，因为正面花萼仅占据很小的图像区域，且容易受到背景干扰、光照变化以及深度特征提取过程中特征衰减的影响。为解决这些挑战，本研究提出了MFE-DETR，一种基于改进RT-DETR的实时检测模型，用于密集遮挡蓝莓和微小花萼的检测。引入微特征增强注意力（MFEA）模块，通过结合异构局部感受野与空间先验偏置来增强弱花萼特征的表示。所提出的模块旨在增强微小花萼区域与周围果实上下文之间的关联，同时保持RT-DETR的端到端检测框架。在自建蓝莓图像数据集上的实验结果表明，MFE-DETR的总体mAP@0.5达到94.2%，正面花萼AP@0.5为91.5%。此外，在NVIDIA Jetson AGX Orin平台上的部署实验表明，所提出的模型在测试配置下实现了实时边缘端推理，表明其作为后续机器人蓝莓采摘系统视觉感知模块的潜力。",null,"Computers and Electronics in Agriculture","2026-09-12T00:00:00Z","论文",10,false,79,{"impact":17,"substance":18,"depth":19,"authority":20,"freshness":21,"relevant":22,"comment":23},16,22,18,14,9,1,"提出MFE-DETR改进模型实现密集遮挡下蓝莓及微小花萼的实时检测，mAP@0.5达94.2%并在Jetson边缘平台验证，对采摘机器人视觉感知有实质参考价值。",[25],{"name":10,"url":6},[27,28,29,30,31],"智慧农业","农业人工智能","目标检测","采摘机器人","蓝莓",[],{"status":34,"locked":14,"version":35,"method":36,"evidence":37},"needs_review","2026-10-22.2","rules",{"candidates":38,"agricultural_evidence":14,"reason":60},[39,45,50,55],{"code":40,"score":41,"supported":14,"title":42,"abstract":43,"tags":44},"I2",2,[],[],[28,29],{"code":46,"score":22,"supported":14,"title":47,"abstract":48,"tags":49},"B2",[],[],[31],{"code":51,"score":22,"supported":14,"title":52,"abstract":53,"tags":54},"F6",[],[],[30],{"code":56,"score":22,"supported":14,"title":57,"abstract":58,"tags":59},"I1",[],[],[27],"农业相关性需要复核",[62,63],"农业人工智能 采摘机器人 智慧农业 目标检测","农业人工智能 采摘机器人","农业人工智能采摘机器人智慧农业目标检测-2277",5,"10.1016\u002Fj.compag.2026.112424",{"doi":66,"openalex_id":68,"authors":69,"venue":10,"cited_by_count":77,"oa_url":6,"card":78,"direction":82,"ingested_from":84},"W7212359038",[70,72,75],{"name":71,"orcid":9},"Sen Li",{"name":73,"orcid":74},"Xiaobin Li","https:\u002F\u002Forcid.org\u002F0000-0002-1099-6046",{"name":76,"orcid":9},"Zhumei Wang",0,{"tldr":79,"method":80,"finding":81,"direction":82,"opportunity":83},"提出改进RT-DETR的MFE-DETR模型，实现密集遮挡下蓝莓及微小花萼的实时检测。","在RT-DETR中引入微特征增强注意力模块MFEA，结合异构局部感受野与空间先验","自建数据集上mAP@0.5达94.2%，前花萼AP@0.5为91.5%，Jetson AGX Ori","农业人工智能与决策模型","可探索微小花萼特征增强模块向其他小目标果实器官检测的迁移，及采摘机器人视觉-控制闭环集成。","openalex","2026-09-13T23:30:01.494783Z",{"total":87,"page":22,"page_size":87,"items":88},6,[89,137,187,241,316,374],{"id":90,"title":91,"url":92,"summary":93,"summary_zh":9,"content":9,"source_name":94,"source_url":9,"published_at":95,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":15,"score_detail":96,"sources":100,"tags":102,"primary_topic":9,"secondary_topics":104,"classification":105,"search_phrases":123,"slug":126,"view_count":127,"doi":9,"paper":128,"created_at":136},4285,"Lightweight AI Gives Harvesting Robots a Sharper Eye for Green Jujubes","https:\u002F\u002Fscienmag.com\u002Flightweight-ai-gives-harvesting-robots-a-sharper-eye-for-green-jujubes","广西师范大学朱红燕团队与浙江大学何勇合作，研发协作感知框架用于青枣采摘机器人。框架包含两个组件：RDA-YOLO是基于YOLOv11n-POSE构建的轻量级神经网络，引入DSConv降低计算开销，设计C3k2_RepViT改善多尺度特征融合，自定义ARC-PoseHead采用对数尺度极坐标回归和亚像素细化提升关键点定位精度；SAM2用于检测框内的细粒度轮廓分割，评估遮挡状态并确定收获优先级。","BMC Plant Biology 2026-10-03 Zhu Hongyan et al. Guangxi Normal Univ & Zhejiang Univ","2026-10-03T00:00:00Z",{"impact":19,"substance":97,"depth":19,"authority":20,"freshness":98,"relevant":22,"comment":99},21,8,"轻量化姿态估计与分割协作框架提升青枣采摘机器人遮挡识别与优先级判断，方法新颖、来源核心期刊，具细分领域参考价值。",[101],{"name":94,"url":92},[27,28,29,30,103],"青枣",[],{"status":34,"locked":14,"version":35,"method":36,"evidence":106},{"candidates":107,"agricultural_evidence":14,"reason":60},[108,114,119],{"code":40,"score":65,"supported":109,"title":110,"abstract":111,"tags":113},true,[],[112],"YOLO",[28,29],{"code":51,"score":115,"supported":109,"title":116,"abstract":117,"tags":118},4,[],[30],[30],{"code":56,"score":22,"supported":14,"title":120,"abstract":121,"tags":122},[],[],[27],[124,125],"广西师范大学 青枣 采摘机器人","RDA-YOLO 青枣 检测","广西师范大学青枣采摘机器人-4285",3,{"doi":9,"openalex_id":9,"authors":129,"venue":9,"cited_by_count":77,"oa_url":9,"card":130,"direction":82,"ingested_from":135},[],{"tldr":131,"method":132,"finding":133,"direction":82,"opportunity":134},"研发协作感知框架，让青枣采摘机器人精准识别与定位果实。","基于YOLOv11n-POSE的RDA-YOLO轻量网络结合SAM2分割，优化关","框架能有效评估遮挡并确定收获优先级，提升采摘机器人感知精度。","可探索轻量模型在复杂果园环境下的泛化能力及多机器人协同采摘决策。","agent","2026-10-06T00:05:04.907455Z",{"id":138,"title":139,"url":140,"summary":141,"summary_zh":9,"content":9,"source_name":142,"source_url":9,"published_at":143,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":144,"score_detail":145,"sources":151,"tags":153,"primary_topic":9,"secondary_topics":155,"classification":156,"search_phrases":175,"slug":178,"view_count":87,"doi":9,"paper":179,"created_at":186},4158,"基于改进YOLOv8的火龙果采摘机器人系统设计与研究","https:\u002F\u002Fm.chinaaet.com\u002Farticle\u002F3000180728","刘爱华、支铁山（电子技术应用\u002F集成电路应用2026-09-28；广西物流职业技术学院、广西自然资源职业技术学院）。提出针对火龙果采摘的改进YOLOv8算法：1）在主干网络中引入可变形卷积增强对不规则形状果实的特征提取；2）设计高效多尺度注意力模块强化遮挡果物的上下文建模；3）优化检测头结构提升边界框回归精度。同时设计轻量化采摘机械臂与柔性末端执行器，构建完整的机器人系统并开展田间验证（深度信息计算采摘点三维坐标、电液比例升降平台调整高度、六自由度机械臂移动）。","电子技术应用·2026-09-28","2026-09-28T00:00:00Z",68,{"impact":146,"substance":147,"depth":148,"authority":149,"freshness":87,"relevant":22,"comment":150},12,20,17,13,"论文提出改进YOLOv8检测算法并构建火龙果采摘机器人系统，方法有创新且经田间验证，但属细分作物技术进展，公共影响有限。",[152],{"name":142,"url":140},[27,28,29,30,154],"火龙果",[],{"status":34,"locked":14,"version":35,"method":36,"evidence":157},{"candidates":158,"agricultural_evidence":14,"reason":60},[159,163,167,171],{"code":51,"score":21,"supported":109,"title":160,"abstract":161,"tags":162},[30],[],[30],{"code":40,"score":41,"supported":14,"title":164,"abstract":165,"tags":166},[],[],[28,29],{"code":46,"score":22,"supported":14,"title":168,"abstract":169,"tags":170},[],[],[154],{"code":56,"score":22,"supported":14,"title":172,"abstract":173,"tags":174},[],[],[27],[176,177],"火龙果采摘机器人 YOLOv8","广西物流职业技术学院 火龙果","火龙果采摘机器人YOLOv8-4158",{"doi":9,"openalex_id":9,"authors":180,"venue":9,"cited_by_count":77,"oa_url":9,"card":181,"direction":82,"ingested_from":135},[],{"tldr":182,"method":183,"finding":184,"direction":82,"opportunity":185},"提出改进YOLOv8火龙果检测算法并集成六自由度采摘机器人系统，完成田间验证。","可变形卷积、多尺度注意力、优化检测头，结合深度相机与柔性末端执行器。","改进算法提升不规则与遮挡果实检测精度，机器人系统实现采摘点定位与自动采摘。","可探索多果实遮挡下的实时检测与采摘路径规划协同优化，并拓展至其他藤本作物。","2026-10-04T00:09:10.086989Z",{"id":188,"title":189,"url":190,"summary":191,"summary_zh":9,"content":9,"source_name":192,"source_url":9,"published_at":193,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":15,"score_detail":194,"sources":196,"tags":198,"primary_topic":200,"secondary_topics":206,"classification":207,"search_phrases":229,"slug":232,"view_count":115,"doi":9,"paper":233,"created_at":240},4156,"基于改进YOLOv8n的番茄果实检测模型","http:\u002F\u002Fwww.qikanvip.com\u002Fqkml\u002F174644.html","李慧琴、宋赵铭、肖亚涛、兰明明（河南农业大学学报2026年第02期；河南农业大学机电工程学院）。提出一种基于改进YOLOv8n的番茄果实检测模型，将CBAM注意力机制加入YOLOv8n模型的特征提取环节，引入BiFPN模块实现多尺度特征融合替换原PANet结构，结合损失函数S_I优化训练过程。结果：查准率P=95.3%、召回率R=92.1%、F1分数=93.7%、mAP@0.5=96.7%；检测速度最快1.5ms\u002F目标、最慢不超过19.6ms，可应用于番茄采摘机器人提升智能水平。","河南农业大学学报2026年第02期","2026-10-01T00:00:00Z",{"impact":17,"substance":18,"depth":19,"authority":20,"freshness":21,"relevant":22,"comment":195},"核心期刊论文，方法有改进且指标扎实，对采摘机器人视觉识别有实用价值，值得入选。",[197],{"name":192,"url":190},[27,28,29,30,199],"番茄检测",{"code":201,"slug":202,"name":203,"path":204,"parent_code":205},"B1","vegetables","蔬菜","\u002Ftopic\u002Fhorticulture\u002Fvegetables","B",[],{"status":208,"locked":14,"version":35,"method":36,"evidence":209},"classified",{"candidates":210,"agricultural_evidence":109},[211,217,221,225],{"code":201,"score":212,"supported":109,"title":213,"abstract":215,"tags":216},11,[214],"番茄",[214],[],{"code":51,"score":115,"supported":109,"title":218,"abstract":219,"tags":220},[],[30],[30],{"code":40,"score":41,"supported":14,"title":222,"abstract":223,"tags":224},[],[],[28,29],{"code":56,"score":22,"supported":14,"title":226,"abstract":227,"tags":228},[],[],[27],[230,231],"河南农业大学 番茄 采摘机器人","改进YOLOv8n 番茄检测","河南农业大学番茄采摘机器人-4156",{"doi":9,"openalex_id":9,"authors":234,"venue":9,"cited_by_count":77,"oa_url":9,"card":235,"direction":82,"ingested_from":135},[],{"tldr":236,"method":237,"finding":238,"direction":82,"opportunity":239},"提出改进YOLOv8n番茄果实检测模型，提升采摘机器人识别精度与速度。","在YOLOv8n中引入CBAM注意力、BiFPN多尺度融合并优化损失函数S_I。","查准率95.3%、mAP@0.5达96.7%，单目标检测最快1.5ms。","可探索复杂光照与遮挡场景下的轻量化模型，并融合三维定位实现采摘决策闭环。","2026-10-04T00:09:09.963710Z",{"id":242,"title":243,"url":244,"summary":245,"summary_zh":246,"content":9,"source_name":247,"source_url":244,"published_at":95,"category":12,"cover_url":9,"hotness":13,"is_selected":109,"score":248,"score_detail":249,"sources":251,"tags":253,"primary_topic":254,"secondary_topics":259,"classification":260,"search_phrases":283,"slug":286,"view_count":115,"doi":287,"paper":288,"created_at":315},4055,"A collaborative perception framework for green jujube: from lightweight key point detection to posture-aware harvest analysis","https:\u002F\u002Fdoi.org\u002F10.1186\u002Fs12870-026-09842-7","Abstract Automated harvesting is pivotal to smart agriculture, yet precisely localizing picking points in complex orchard environments remains a key challenge for robotic harvesting systems. To this end, this study proposes a collaborative perception framework integrating a lightweight pose estimation model termed RepViT-DSConv-ARC-PoseHead enhanced YOLO (RDA-YOLO) with Segment Anything Model 2 (SAM2). RDA-YOLO is the lightweight component responsible for real-time fruit detection and keypoint localization, while SAM2 is adopted for fruit contour segmentation and occlusion assessment. Based on the YOLOv11n-POSE architecture, we introduce Distribution Shift Convolution (DSConv) to reduce computational overhead and design the C3k2_RepViT module to enhance multi-scale feature fusion. Crucially, a specialized Adaptive Regression Coordination Pose Head (ARC-PoseHead) is developed to replace traditional prediction structures, employing log-scale polar regression and subpixel refinement to significantly improve coordinate perception of jujubes and key points under complex backgrounds. RDA-YOLO performs real-time detection and locates key points such as the pedicel-fruit transition zone to calculate optimal harvesting angles, while SAM2 performs fine-grained contour segmentation based on detection boxes to determine occlusion status and establish harvesting priority. Experimental results show that RDA-YOLO achieves 93.9% precision and 94.9% mAP@50 for green jujube detection, while for key point detection, it attains a superior precision of 90.4% and a mAP@50 of 83.6%. Notably, the model achieves a lightweight design by reducing parameters to 2.490 million and computational load to 5.7 GFLOPs. Deployment feasibility was verified on the field robot control computer and a robotic platform using jujube fruit models, demonstrating the potential of the proposed framework for practical harvesting applications.","自动采摘是智慧农业的关键环节，但在复杂果园环境中精准定位采摘点仍是机器人采摘系统面临的核心挑战。为此，本研究提出了一种协同感知框架，将轻量级姿态估计模型RDA-YOLO（RepViT-DSConv-ARC-PoseHead增强型YOLO）与Segment Anything Model 2（SAM2）相结合。RDA-YOLO作为轻量级组件，负责实时果实检测与关键点定位；SAM2则用于果实轮廓分割与遮挡评估。基于YOLOv11n-POSE架构，我们引入分布偏移卷积（DSConv）以降低计算开销，并设计C3k2_RepViT模块以增强多尺度特征融合。关键在于，我们开发了专用的自适应回归协调姿态头（ARC-PoseHead）以替代传统预测结构，采用对数尺度极坐标回归和亚像素细化，显著提升了复杂背景下枣果及关键点的坐标感知能力。RDA-YOLO执行实时检测并定位果柄-果实过渡区等关键点以计算最优采摘角度，而SAM2基于检测框进行细粒度轮廓分割，以判定遮挡状态并建立采摘优先级。实验结果表明，RDA-YOLO对青枣检测的精确率达到93.9%，mAP@50为94.9%；在关键点检测方面，精确率达到90.4%，mAP@50为83.6%。值得注意的是，该模型通过将参数量降至249万、计算量降至5.7 GFLOPs，实现了轻量化设计。在田间机器人控制计算机及使用枣果模型的机器人平台上验证了部署可行性，展示了所提框架在实际采摘应用中的潜力。","BMC Plant Biology",82,{"impact":19,"substance":18,"depth":19,"authority":20,"freshness":13,"relevant":22,"comment":250},"提出轻量化姿态估计与SAM2协同感知框架，实现青枣采摘点定位与遮挡评估，方法新颖、指标扎实，对智能采收有实用价值。",[252],{"name":247,"url":244},[27,28,29,30,103],{"code":40,"slug":255,"name":256,"path":257,"parent_code":258},"agricultural-ai","农业AI与大模型","\u002Ftopic\u002Fdigital-agriculture\u002Fagricultural-ai","I",[],{"status":208,"locked":14,"version":35,"method":36,"evidence":261},{"candidates":262,"agricultural_evidence":109},[263,269,273,278],{"code":56,"score":264,"supported":109,"title":265,"abstract":266,"tags":268},7,[],[27,267],"smart agriculture",[27],{"code":40,"score":65,"supported":109,"title":270,"abstract":271,"tags":272},[],[112],[28,29],{"code":51,"score":115,"supported":109,"title":274,"abstract":275,"tags":277},[],[276],"robotic harvesting",[30],{"code":46,"score":127,"supported":109,"title":279,"abstract":280,"tags":282},[],[281],"orchard",[],[284,285],"RDA-YOLO SAM2 青枣 采摘","青枣 关键点检测 采摘机器人","RDA-YOLOSAM2青枣采摘-4055","10.1186\u002Fs12870-026-09842-7",{"doi":287,"openalex_id":289,"authors":290,"venue":247,"cited_by_count":77,"oa_url":244,"card":309,"direction":314,"ingested_from":84},"W7216260984",[291,294,297,299,301,303,305,307],{"name":292,"orcid":293},"Hongyan Zhu","https:\u002F\u002Forcid.org\u002F0000-0002-8703-6289",{"name":295,"orcid":296},"Runlong Cao","https:\u002F\u002Forcid.org\u002F0009-0002-9535-4946",{"name":298,"orcid":9},"Shuai Qin",{"name":300,"orcid":9},"Chengzhi Lin",{"name":302,"orcid":9},"Weizhe Sun",{"name":304,"orcid":9},"Yanli Zou",{"name":306,"orcid":9},"Zhixian Liao",{"name":308,"orcid":9},"Yong He",{"tldr":310,"method":311,"finding":312,"direction":82,"opportunity":313},"提出RDA-YOLO与SAM2协同感知框架，实现青枣采摘点定位与姿态分析。","改进YOLOv11n-POSE，引入DSConv、RepViT与ARC-Pose","检测精度93.9%、关键点精度90.4%，模型仅2.49M参数、5.7 GFLOPs。","可探索多作物通用轻量姿态估计与遮挡感知采摘决策的迁移研究。","智慧农业 \u002F 农业物联网","2026-10-03T23:30:09.616335Z",{"id":317,"title":318,"url":319,"summary":320,"summary_zh":9,"content":9,"source_name":321,"source_url":9,"published_at":322,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":323,"score_detail":324,"sources":327,"tags":329,"primary_topic":331,"secondary_topics":335,"classification":336,"search_phrases":362,"slug":365,"view_count":127,"doi":9,"paper":366,"created_at":373},2490,"Lightweight Detection of Blueberries at Different Maturity Stages in Complex Orchard Environments(YOLOv12n+SHSA2C2f+知识蒸馏)","https:\u002F\u002Fwww.mdpi.com\u002F2077-0472\u002F16\u002F18\u002F1949","云南农业大学大数据学院联合云南省农业大数据工程技术研究中心 2026 年 9 月 10 日在《Agriculture》发表。针对小果、密集、遮挡、色彩过渡细微等复杂田间条件下的蓝莓成熟期检测难题,构建 4680 图像 60573 标注框田间数据集,重新设计 YOLOv12n 的 P3\u002FP4\u002FP5 检测结构为 P2\u002FP3\u002FP4 架构,引入轻量化 SHSA2C2f 单头自注意力特征增强模块,DINOv3 (ViT-S\u002F16) 引导的非对称知识蒸馏。精度导向 M4 模型 mAP@0.5 达 92.68%±0.04%、mAP@0.5:0.95 86.38%±0.10%,参数量仅 0.79 M;经 TensorRT FP16 转换后保留 92.75% mAP@0.5,可在 Jetson Orin Nano 上以 15.9 FPS 运行。","MDPI Agriculture 2026-09-10","2026-09-09T16:00:00Z",80,{"impact":19,"substance":325,"depth":19,"authority":149,"freshness":98,"relevant":22,"comment":326},23,"面向复杂果园的蓝莓成熟期轻量化检测研究，方法新颖、数据规模扎实且给出边缘部署实测性能，具备较高专业参考价值。",[328],{"name":321,"url":319},[27,28,330,29,31],"边缘计算",{"code":46,"slug":332,"name":333,"path":334,"parent_code":205},"fruit","果树","\u002Ftopic\u002Fhorticulture\u002Ffruit",[],{"status":208,"locked":14,"version":35,"method":36,"evidence":337},{"candidates":338,"agricultural_evidence":109},[339,343,349,353,357],{"code":46,"score":21,"supported":109,"title":340,"abstract":341,"tags":342},[281],[],[31],{"code":344,"score":127,"supported":109,"title":345,"abstract":346,"tags":348},"I6",[],[347],"农业大数据",[],{"code":40,"score":41,"supported":14,"title":350,"abstract":351,"tags":352},[],[],[28,29],{"code":56,"score":22,"supported":14,"title":354,"abstract":355,"tags":356},[],[],[27],{"code":358,"score":22,"supported":14,"title":359,"abstract":360,"tags":361},"I5",[],[],[330],[363,364],"农业人工智能 智慧农业 目标检测 边缘计算","农业人工智能 智慧农业","农业人工智能智慧农业目标检测边缘计算-2490",{"doi":9,"openalex_id":9,"authors":367,"venue":9,"cited_by_count":77,"oa_url":9,"card":368,"direction":82,"ingested_from":135},[],{"tldr":369,"method":370,"finding":371,"direction":82,"opportunity":372},"提出轻量YOLOv12n改进模型，实现复杂果园中多成熟期蓝莓的实时检测。","4680张田间图像数据集，P2\u002FP3\u002FP4结构、SHSA2C2f模块与DINOv","M4模型mAP@0.5达92.68%，仅0.79M参数，Jetson Orin Nano上15.9 ","可探索多作物通用轻量检测框架，或结合成熟度计数用于产量预测与采摘决策。","2026-09-15T00:04:27.061523Z",{"id":375,"title":376,"url":377,"summary":378,"summary_zh":9,"content":9,"source_name":379,"source_url":9,"published_at":380,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":381,"score_detail":382,"sources":384,"tags":386,"primary_topic":389,"secondary_topics":395,"classification":397,"search_phrases":420,"slug":423,"view_count":77,"doi":9,"paper":424,"created_at":432},5532,"精准棉花病害管理：融合无人机影像与定制YOLO深度学习框架","https:\u002F\u002Farxiv.org\u002Fabs\u002F2610.24182","研究提出单阶段全卷积检测器Cotton-Vision，融合幽灵瓶颈通道注意力与跨阶段部分融合三项轻量化设计，参数减少20%。在巴基斯坦旁遮普省7个棉田采集的2700余幅高分辨率无人机影像上测试，精确率97.7%、召回率77.7%、mAP@50为85.6%，较主流基线提升10至20个百分点。预印本，仅在巴基斯坦单一区域验证。","arXiv 预印本 2026-10-16","2026-10-16T00:00:00Z",69,{"impact":17,"substance":97,"depth":148,"authority":21,"freshness":87,"relevant":22,"comment":383},"方法轻量化与实测数据有增量，但仅单区域预印本验证，属细分技术进展，可入精选备选。",[385],{"name":379,"url":377},[27,28,29,387,388],"无人机遥感","棉花病害",{"code":390,"slug":391,"name":392,"path":393,"parent_code":394},"A6","cash-crops","棉麻糖与特色经作","\u002Ftopic\u002Ffield-crops\u002Fcash-crops","A",[396],{"code":40,"slug":255,"name":256,"path":257,"parent_code":258},{"status":208,"locked":14,"version":35,"method":36,"evidence":398},{"candidates":399,"agricultural_evidence":109},[400,405,411,415],{"code":40,"score":19,"supported":109,"title":401,"abstract":403,"tags":404},[402,112],"深度学习",[],[28,29],{"code":390,"score":212,"supported":109,"title":406,"abstract":408,"tags":410},[407],"棉花",[409],"cotton",[],{"code":56,"score":22,"supported":14,"title":412,"abstract":413,"tags":414},[],[],[27],{"code":416,"score":22,"supported":14,"title":417,"abstract":418,"tags":419},"I3",[],[],[387],[421,422],"Cotton-Vision 棉花病害","无人机影像 棉花 检测","Cotton-Vision棉花病害-5532",{"doi":9,"openalex_id":9,"authors":425,"venue":9,"cited_by_count":77,"oa_url":9,"card":426,"direction":430,"ingested_from":135},[],{"tldr":427,"method":428,"finding":429,"direction":430,"opportunity":431},"提出轻量化检测器Cotton-Vision，用无人机影像精准识别棉花病害。","基于YOLO改进，融合幽灵瓶颈通道注意力与跨阶段部分融合，用2700余幅无人机影","精确率97.7%、mAP@50达85.6%，较主流基线提升10至20个百分点。","农业遥感与作物表型","仅在巴基斯坦单一区域验证，可扩展至多区域多品种，并探索边缘设备实时部署。","2026-10-22T00:07:09.816790Z"]