[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"item-2934":3,"related-2934":52},{"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":27,"search_phrases":32,"slug":35,"view_count":36,"doi":37,"paper":38,"created_at":51},2934,"A CNN-Based Approach for Leaf Disease Prediction in Smart Agriculture","https:\u002F\u002Fdoi.org\u002F10.5281\u002Fzenodo.22825329","Plants play a crucial role in sustaining life by serving as a primary source of energy and mitigating global warming. However, they are increasingly vulnerable to diseases such as bacterial spot, late blight, and Septoria leaf spot, which significantly impact crop yield and agricultural productivity. Early and accurate detection of these diseases is essential for effective disease management and improved agricultural outcomes. This project aims to develop a deep learning-based approach for detecting plant leaf diseases using Convolutional Neural Networks (CNN). By leveraging benchmark datasets, the proposed CNN model demonstrates superior performance compared to traditional machine learning techniques, achieving an accuracy of 92%, precision of 89%, F1-score of 93%, and recall of 92.47%. The results highlight the effectiveness of CNN in automating disease identification, enabling timely intervention, and promoting sustainable agricultural practices.","植物在维持生命方面发挥着至关重要的作用，既是主要的能量来源，又能缓解全球变暖。然而，植物日益受到细菌性斑点病、晚疫病和壳针孢叶斑病等病害的威胁，严重影响作物产量和农业生产率。早期准确地检测这些病害对于有效防控病害和改善农业成果至关重要。本项目旨在开发一种基于深度学习的方法，利用卷积神经网络（CNN）检测植物叶片病害。通过利用基准数据集，所提出的CNN模型展现出优于传统机器学习技术的性能，达到了92%的准确率、89%的精确率、93%的F1分数和92.47%的召回率。结果表明，CNN在自动化病害识别方面具有显著效果，能够实现及时干预并促进可持续农业实践。",null,"Zenodo (CERN European Organization for Nuclear Research)","2026-09-18T00:00:00Z","论文",25,false,65,{"impact":17,"substance":18,"depth":19,"authority":17,"freshness":20,"relevant":21,"comment":22},12,18,14,9,1,"基于CNN的叶片病害识别研究，方法常规、数据集为公开基准，准确率92%属中等水平，对智慧农业植保场景有一定参考价值但缺乏突破性。",[24,25],{"name":10,"url":6},{"name":10,"url":26},"https:\u002F\u002Fdoi.org\u002F10.5281\u002Fzenodo.22825330",[28,29,30,31],"智慧农业","农业人工智能","深度学习","植物病害识别",[33,34],"CNN 植物叶片病害 识别","卷积神经网络 作物病害 检测","CNN植物叶片病害识别-2934",0,"10.5281\u002Fzenodo.22825329",{"doi":37,"openalex_id":39,"authors":40,"venue":10,"cited_by_count":36,"oa_url":6,"card":43,"direction":49,"ingested_from":50},"W7213587327",[41],{"name":42,"orcid":9},"B.Yashmal Sai, K.Karthik, K.Neeraj, G. Mahabub Subhani",{"tldr":44,"method":45,"finding":46,"direction":47,"opportunity":48},"用CNN对植物叶片病害进行自动识别，在基准数据集上取得92%准确率。","基于卷积神经网络，使用植物叶片病害基准数据集训练与评估。","CNN优于传统机器学习方法，准确率92%、F1值93%，可支持及时干预。","农业人工智能与决策模型","可探索轻量化CNN在田间移动端实时检测，并结合多病害与早期症状识别。","智慧农业 \u002F 农业物联网","openalex","2026-09-19T23:30:11.855376Z",{"total":53,"page":21,"page_size":53,"items":54},6,[55,95,134,175,207,246],{"id":56,"title":57,"url":58,"summary":59,"summary_zh":60,"content":9,"source_name":61,"source_url":58,"published_at":62,"category":12,"cover_url":9,"hotness":63,"is_selected":14,"score":64,"score_detail":65,"sources":70,"tags":72,"search_phrases":73,"slug":76,"view_count":36,"doi":77,"paper":78,"created_at":94},1617,"MRC-Net: a reliable plant disease classification framework with multi-frequency state-space enhancement and conformal prediction","https:\u002F\u002Fdoi.org\u002F10.3389\u002Ffpls.2026.1895287","Plant disease image classification is a key task in disease monitoring and precise prevention and control in smart agriculture. However, complex backgrounds, finegrained lesion differences, and model overconfidence still limit recognition performance and practical application reliability. To address these issues, this paper proposes a reliable plant disease recognition framework that integrates multi-frequency selective state-space feature enhancement with conformal-aware reliable prediction. The proposed method adopts Swin-Tiny as the backbone network and uses the MF-SSFE module to jointly model low-frequency leaf structures, high-frequency lesion textures, and cross-region state-space contexts, thereby enhancing disease-related evidence in complex scenarios. Meanwhile, the CARP module is introduced to incorporate the idea of conformal prediction into the classification output process, enabling the model to express uncertainty while providing class predictions. Experimental results show that the proposed method achieves Accuracy values of 0.9891, 0.3889, and 0.9800 on the NGLD, PlantDoc, and PlantVillage datasets, respectively, and AUC values of 0.9995, 0.8690, and 0.9989, respectively, outperforming the comparison methods. Ablation experiments and sensitivity analysis further verify the effectiveness and stability of each module. Model complexity analysis shows that the proposed method maintains acceptable computational overhead while achieving superior recognition performance.","植物病害图像分类是智慧农业中病害监测与精准防控的关键任务。然而，复杂背景、细粒度病斑差异以及模型过度自信等问题仍制约着识别性能与实际应用的可靠性。针对上述问题，本文提出了一种融合多频选择性状态空间特征增强与一致性感知可靠预测的可靠植物病害识别框架。该方法以Swin-Tiny为骨干网络，利用MF-SSFE模块联合建模低频叶片结构、高频病斑纹理及跨区域状态空间上下文，从而增强复杂场景下与病害相关的证据信息。同时，引入CARP模块，将一致性预测的思想融入分类输出过程，使模型在提供类别预测的同时能够表达不确定性。实验结果表明，该方法在NGLD、PlantDoc和PlantVillage数据集上的准确率分别达到0.9891、0.3889和0.9800，AUC值分别达到0.9995、0.8690和0.9989，优于对比方法。消融实验和敏感性分析进一步验证了各模块的有效性与稳定性。模型复杂度分析表明，该方法在保持可接受计算开销的同时，实现了优越的识别性能。","Frontiers in Plant Science","2026-09-03T00:00:00Z",10,70,{"impact":17,"substance":66,"depth":18,"authority":67,"freshness":68,"relevant":21,"comment":69},20,13,7,"提出结合多频状态空间与保形预测的植物病害分类框架，在多个数据集上表现优异，兼具可靠性与效率。",[71],{"name":61,"url":58},[28,29,30,31],[74,75],"农业人工智能 植物病害识别 智慧农业 深度学习","农业人工智能 植物病害识别","农业人工智能植物病害识别智慧农业深度学习-1617","10.3389\u002Ffpls.2026.1895287",{"doi":77,"openalex_id":79,"authors":80,"venue":61,"cited_by_count":36,"oa_url":58,"card":89,"direction":49,"ingested_from":50},"W7207549269",[81,83,86],{"name":82,"orcid":9},"Shiyao Xie",{"name":84,"orcid":85},"X. H. Zhang","https:\u002F\u002Forcid.org\u002F0009-0003-9957-0002",{"name":87,"orcid":88},"Yang Li","https:\u002F\u002Forcid.org\u002F0000-0002-3006-7420",{"tldr":90,"method":91,"finding":92,"direction":47,"opportunity":93},"提出可靠植物病害分类框架，结合多频状态空间增强与保形预测，提升复杂场景识别精度与可靠性。","Swin-Tiny骨干，MF-SSFE多频特征增强，CARP保形预测模块，在NG","在三个数据集上准确率分别达0.9891、0.3889、0.9800，AUC达0.9995、0.869","可探索将保形预测用于其他农业任务（如产量预测）以量化不确定性，或优化多频特征提取以应对更复杂田间场景。","2026-09-04T23:30:09.854906Z",{"id":96,"title":97,"url":98,"summary":99,"summary_zh":100,"content":9,"source_name":101,"source_url":98,"published_at":11,"category":12,"cover_url":9,"hotness":63,"is_selected":14,"score":102,"score_detail":103,"sources":107,"tags":109,"search_phrases":112,"slug":115,"view_count":36,"doi":116,"paper":117,"created_at":133},2963,"A comprehensive review of deep learning methods for weed classification in precision agriculture","https:\u002F\u002Fdoi.org\u002F10.1007\u002Fs44163-026-01916-7","Weeds today are among the factors contributing to low agricultural productivity. As the world’s population continues to grow, there is an urgent need to meet global food demand. Nigeria currently lacks sufficient crop production to feed its growing population, and weeds are among the core contributors to poor agricultural yield. This study conducts a comprehensive review of Deep Learning (DL) approaches to weed classification in precision agriculture, covering literature from 2018 to 2025, was carried out. We employed a mix of quantitative and qualitative methods in the course of this review paper. Our data source is centred on Scopus-indexed papers, published with Sensors, Electronics, and Agriculture in MDPI as well as IEEE, Thomson Reuters, and Springer. The study systematically reviewed and analysed machine learning (ML), DL, and instance segmentation techniques to identify the key technological and environmental barriers, such as data limitations, class imbalance, environmental variability, and model scalability issues that affect the effectiveness and efficiency of these models when deployed in real time. These findings show that while weed management models like the YOLO variants, ResNet, and Vision Transformers achieved high accuracy in training and testing, they are associated with several challenges in their real world-deployment, such as occlusion, small object detection, and environmental adaptability. Overall, this research provides recommended solutions to enhance model robustness, scalability, and efficiency. It further provides a summary of the current state and future directions for AI-driven weed management.","杂草是当前导致农业生产力低下的因素之一。随着世界人口持续增长，满足全球粮食需求已成为迫切任务。尼日利亚目前的作物产量不足以养活其不断增长的人口，而杂草是导致农业产量低下的核心因素之一。本研究对精准农业中基于深度学习（Deep Learning，DL）的杂草分类方法进行了全面综述，涵盖2018年至2025年的文献。在综述过程中，我们采用了定量与定性相结合的方法。数据来源集中于Scopus索引论文，这些论文发表于MDPI旗下的Sensors、Electronics和Agriculture，以及IEEE、Thomson Reuters和Springer。本研究系统综述并分析了机器学习（Machine Learning，ML）、深度学习及实例分割技术，以识别影响这些模型实时部署效果与效率的关键技术和环境障碍，如数据局限性、类别不平衡、环境变异性及模型可扩展性问题。研究结果表明，尽管YOLO系列、ResNet和视觉Transformer（Vision Transformer）等杂草管理模型在训练和测试中达到了较高精度，但在实际部署中仍面临诸多挑战，如遮挡、小目标检测和环境适应性等问题。总体而言，本研究提出了增强模型鲁棒性、可扩展性和效率的推荐解决方案，并进一步总结了人工智能驱动杂草管理的现状与未来方向。","Discover Artificial Intelligence",67,{"impact":17,"substance":18,"depth":104,"authority":67,"freshness":105,"relevant":21,"comment":106},16,8,"系统综述2018—2025年深度学习杂草分类方法，指出遮挡、小目标与环境适应性等落地瓶颈，对农业AI研究有参考价值，但属综述类论文、非突破性成果。",[108],{"name":101,"url":98},[28,29,30,110,111],"杂草识别","精准农业",[113,114],"深度学习 杂草分类 精准农业","YOLO 杂草识别 模型部署","深度学习杂草分类精准农业-2963","10.1007\u002Fs44163-026-01916-7",{"doi":116,"openalex_id":118,"authors":119,"venue":101,"cited_by_count":36,"oa_url":98,"card":128,"direction":47,"ingested_from":50},"W7213558223",[120,122,124,126],{"name":121,"orcid":9},"Njoku Camillus Ekene",{"name":123,"orcid":9},"Francis A. Okoye",{"name":125,"orcid":9},"Ebere Uzoka Chidi",{"name":127,"orcid":9},"OGBU MARY NNENNA",{"tldr":129,"method":130,"finding":131,"direction":47,"opportunity":132},"综述2018-2025年深度学习杂草分类方法，分析技术瓶颈并给出改进建议。","混合定量定性法，基于Scopus及MDPI、IEEE等文献，分析ML、DL与实例","YOLO、ResNet、ViT等精度高，但实际部署受遮挡、小目标与环境适应性限制。","可研究轻量化、跨域自适应模型，解决小目标与遮挡下的实时杂草识别难题。","2026-09-19T23:30:56.875661Z",{"id":135,"title":136,"url":137,"summary":138,"summary_zh":139,"content":9,"source_name":140,"source_url":137,"published_at":141,"category":12,"cover_url":9,"hotness":63,"is_selected":14,"score":142,"score_detail":143,"sources":146,"tags":148,"search_phrases":151,"slug":154,"view_count":36,"doi":155,"paper":156,"created_at":174},2962,"A Multi-Modal Generative Model for Tomato Disease Leaves Understanding","https:\u002F\u002Fdoi.org\u002F10.48550\u002Farxiv.2609.19555","Artificial intelligence for plant disease analysis has advanced from task-specific classifiers to multi-modal models capable of jointly interpreting visual and textual information. However, practical deployment in precision agriculture remains limited because most existing approaches treat disease understanding as isolated prediction tasks, failing to capture the complementary relationships among symptom recognition, severity assessment, and question-driven diagnostic reasoning. In tomato pathology, accurate interpretation of diseased leaves requires more than label prediction; it demands integrating visual symptoms with semantic context to support a comprehensive and explainable understanding. Here, we present SOLAR, a multimodal generative model that understands tomato disease spanning six question-answering tasks. SOLAR learns to align visual features with task-aware language representations by Fusion Expert module based on mixture-of-expert, enabling it to generate contextually relevant answers across diverse diagnostic tasks. By formulating tomato disease analysis as a generative Visual Question Answering (VQA) task, SOLAR provides a flexible framework that supports multi-task inference within a single model while improving performance and cross-task knowledge sharing. We evaluate SOLAR on $41,677$ images, including $216,209$ Question-Answering (QA) pairs to understand tomato leaf disease under both closed and open-ended QA settings. Experimental results show that SOLAR consistently outperforms state-of-the-art vision-only, vision-language, and task-specific models across all tasks, demonstrating superior accuracy, robustness, and multimodal reasoning. These findings highlight the potential of generative multimodal modeling as an effective direction for understanding of plant disease. The code for this study is available at https:\u002F\u002Fgithub.com\u002FEnalisUs\u002FSOLAR.","人工智能用于植物病害分析已从任务专用分类器发展到能够联合解读视觉与文本信息的多模态模型。然而，在精准农业中的实际部署仍然有限，因为大多数现有方法将病害理解视为孤立的预测任务，未能捕捉症状识别、严重程度评估与问题驱动诊断推理之间的互补关系。在番茄病理学中，对病叶的准确解读需要的不仅仅是标签预测；它要求将视觉症状与语义上下文相结合，以支持全面且可解释的理解。在此，我们提出SOLAR，一种多模态生成模型，能够理解番茄病害并涵盖六项问答任务。SOLAR通过基于专家混合的融合专家模块，学习将视觉特征与任务感知的语言表示对齐，使其能够在多样化的诊断任务中生成上下文相关的答案。通过将番茄病害分析表述为生成式视觉问答（VQA）任务，SOLAR提供了一个灵活的框架，支持在单一模型内进行多任务推理，同时提升性能并促进跨任务知识共享。我们在$41,677$张图像上评估SOLAR，其中包括$216,209$个问答（QA）对，以在封闭式和开放式问答设置下理解番茄叶片病害。实验结果表明，SOLAR在所有任务上均持续优于最先进的纯视觉、视觉语言和任务专用模型，展现出更优的准确性、鲁棒性和多模态推理能力。这些发现凸显了生成式多模态建模作为理解植物病害的有效方向的潜力。本研究的代码可在https:\u002F\u002Fgithub.com\u002FEnalisUs\u002FSOLAR获取。","arXiv (Cornell University)","2026-09-17T00:00:00Z",79,{"impact":18,"substance":144,"depth":18,"authority":67,"freshness":105,"relevant":21,"comment":145},22,"提出面向番茄病害的多模态生成式VQA模型SOLAR，在4万余张图像上验证多任务诊断性能，方法新颖、数据规模可观，对智慧植保有参考价值。",[147],{"name":140,"url":137},[28,29,149,31,150],"番茄病害","多模态模型",[152,153],"SOLAR 番茄病害 多模态模型","番茄叶片病害 视觉问答","SOLAR番茄病害多模态模型-2962","10.48550\u002Farxiv.2609.19555",{"doi":155,"openalex_id":157,"authors":158,"venue":140,"cited_by_count":36,"oa_url":137,"card":169,"direction":47,"ingested_from":50},"W7213586905",[159,162,164,166],{"name":160,"orcid":161},"Khang Nguyen Quoc","https:\u002F\u002Forcid.org\u002F0000-0003-4927-4822",{"name":163,"orcid":9},"Minh-Phuoc Tran",{"name":165,"orcid":9},"Gia-Han Truong",{"name":167,"orcid":168},"Luyl-Da Quach","https:\u002F\u002Forcid.org\u002F0000-0002-5661-4250",{"tldr":170,"method":171,"finding":172,"direction":47,"opportunity":173},"提出多模态生成模型SOLAR，统一理解番茄病害叶片的六类问答任务。","基于混合专家融合模块对齐视觉与任务感知语言，用4万余图像和21万问答对训练。","SOLAR在闭集和开放问答中均超越视觉、视觉语言及任务专用模型，展现更强推理能力。","可探索将生成式多模态VQA扩展到更多作物和田间实时场景，并融合传感器数据。","2026-09-19T23:30:55.245800Z",{"id":176,"title":177,"url":178,"summary":179,"summary_zh":180,"content":9,"source_name":181,"source_url":178,"published_at":182,"category":12,"cover_url":9,"hotness":63,"is_selected":14,"score":183,"score_detail":184,"sources":187,"tags":189,"search_phrases":191,"slug":194,"view_count":36,"doi":195,"paper":196,"created_at":206},2940,"A Hybrid CNN–Transformer–LSTM Deep Learning Framework for Automated Cotton Disease Detection","https:\u002F\u002Fdoi.org\u002F10.53365\u002Fnrfhh.1816","Cotton production is frequently affected by leaf diseases that can reduce plant productivity, deteriorate crop quality, and cause considerable financial losses for farmers. Consequently, rapid and reliable disease identification is an important requirement for precision agriculture and effective crop protection. Conventional image-based approaches predominantly employ CNN architectures for visual classification; however, such methods may have limited capability in learning long-range spatial dependencies and modeling changes associated with disease development over time. To overcome these limitations, this research introduces an integrated hybrid deep learning framework that combines Convolutional Neural Networks (CNNs), Transformer-based self-attention, and Long Short-Term Memory (LSTM) networks for intelligent cotton leaf disease recognition. The proposed framework utilizes a pre-trained EfficientNet network to extract discriminative spatial characteristics from cotton leaf images. The extracted representations are subsequently processed through a multi-head self-attention Transformer module, which enables the network to identify relationships between distant and relevant regions of the leaf. An LSTM component is then incorporated to learn sequential dependencies and provide a foundation for analyzing disease evolution and progression. To improve model reliability and generalization, the network is trained using an appropriately organized cotton leaf image dataset together with image augmentation, transfer learning, and fine-tuning techniques. The experimental evaluation indicates that the proposed CNN–Transformer–LSTM architecture provides improved disease classification performance and more effective feature learning compared with conventional CNN-based models. Performance assessment using classification metrics, confusion matrices, and ROC curves demonstrates strong discrimination among the considered cotton disease categories. The combined learning of local visual characteristics, global contextual information, and sequential dependencies provides a comprehensive framework for automated cotton disease assessment. The proposed approach can serve as a scalable artificial intelligence solution for precision agriculture applications. Its architecture also provides opportunities for future integration with IoT-based crop monitoring systems, environmental data acquisition, and disease progression prediction, supporting early warning systems and intelligent crop health management in smart farming environments.","棉花生产常受叶片病害影响，这些病害会降低植株生产力、恶化作物品质，并给农民造成相当大的经济损失。因此，快速可靠的病害识别是精准农业和有效作物保护的重要需求。传统的基于图像的方法主要采用CNN架构进行视觉分类；然而，此类方法在学习长程空间依赖关系以及建模与病害随时间发展相关的变化方面能力有限。为克服这些局限，本研究提出了一种集成混合深度学习框架，将卷积神经网络（CNN）、基于Transformer的自注意力机制和长短期记忆（LSTM）网络相结合，用于智能棉花叶片病害识别。所提出的框架利用预训练的EfficientNet网络从棉花叶片图像中提取具有判别力的空间特征。提取到的表示随后通过多头自注意力Transformer模块进行处理，使网络能够识别叶片中相距较远且相关区域之间的关系。随后引入LSTM组件以学习序列依赖关系，并为分析病害演变和进展提供基础。为提高模型可靠性和泛化能力，网络使用组织良好的棉花叶片图像数据集进行训练，并结合图像增强、迁移学习和微调技术。实验评估表明，与传统的基于CNN的模型相比，所提出的CNN–Transformer–LSTM架构提供了更好的病害分类性能和更有效的特征学习。使用分类指标、混淆矩阵和ROC曲线进行的性能评估表明，该方法在所考虑的棉花病害类别之间具有较强的判别能力。局部视觉特征、全局上下文信息和序列依赖关系的联合学习为自动化棉花病害评估提供了一个全面的框架。所提出的方法可作为精准农业应用中可扩展的人工智能解决方案。其架构也为未来与基于物联网的作物监测系统、环境数据采集和病害进展预测的集成提供了机会，从而支持预警系统和智能","Natural Resources for Human Health","2026-09-16T00:00:00Z",69,{"impact":17,"substance":66,"depth":185,"authority":17,"freshness":105,"relevant":21,"comment":186},17,"提出CNN-Transformer-LSTM混合框架用于棉花叶病识别，方法有创新但属实验室验证阶段，产业影响有限。",[188],{"name":181,"url":178},[28,29,30,111,190],"棉花病害识别",[192,193],"棉花叶病 CNN Transformer LSTM","EfficientNet 棉花病害检测","棉花叶病CNNTransformerLSTM-2940","10.53365\u002Fnrfhh.1816",{"doi":195,"openalex_id":197,"authors":198,"venue":181,"cited_by_count":36,"oa_url":9,"card":201,"direction":49,"ingested_from":50},"W7213544712",[199],{"name":200,"orcid":9},"Prajakta Sunil Gupta",{"tldr":202,"method":203,"finding":204,"direction":47,"opportunity":205},"提出CNN-Transformer-LSTM混合框架，实现棉花叶片病害自动识别。","EfficientNet提取特征，Transformer自注意力与LSTM建模，","混合模型分类性能优于传统CNN，能同时学习局部特征、全局上下文和序列依赖。","可融合IoT环境数据与时间序列，开展病害进展预测和早期预警系统研究。","2026-09-19T23:30:19.558449Z",{"id":208,"title":209,"url":210,"summary":211,"summary_zh":212,"content":9,"source_name":213,"source_url":210,"published_at":11,"category":12,"cover_url":9,"hotness":63,"is_selected":14,"score":214,"score_detail":215,"sources":217,"tags":219,"search_phrases":222,"slug":225,"view_count":36,"doi":226,"paper":227,"created_at":245},2910,"Precision Diagnosis of Apple Leaf Diseases Across Infection Stages via Web-Based Analysis","https:\u002F\u002Fdoi.org\u002F10.1007\u002Fs11119-026-10451-5","Abstract Purpose Throughout the growing season, leaf diseases pose a huge risk to apple fruit yield and quality. While early detection targets timely intervention, understanding disease progression across different stages is essential for effective treatment and resistance breeding. Methods This study proposes a deep learning-based detection system that identifies apple leaf diseases at various infection stages in real-world environments. Most existing methods have focused on disease detection in controlled conditions, while the developed system operates in complex field settings with variable lighting and background noise. Two YOLOv8-based variants integrated with a multi-head self-attention (MHSA) module were developed to improve detection of dense, tiny, and feature-similar lesions. In addition, a web-based interactive tool was built by combining the Segment Anything Model (SAM) with the detection model to support leaf isolation, disease localization, infection estimation, severity categorization, and automatic report generation. Results The training results on proposed PA-ALeaf dataset demonstrated that YOLOv8-MHSA_b_h varaint achieved the highest performance (Precision 60.5%, mAP50 56.3%, F1 confidence 56.0%), while maintaining real-time inference (12.38 ms per image), compared to state-of-the-art models. Conclusion By bridging practical disease monitoring and scientific research, our system offers a comprehensive, scalable solution for apple leaf disease detection in real-world orchards. This system could benefit both farmers by enabling multi-stage intervention and researchers by providing insights into disease progression.","摘要 目的 在整个生长季中，叶片病害对苹果果实产量和品质构成巨大风险。虽然早期检测有助于及时干预，但了解病害在不同阶段的进展对于有效治疗和抗性育种至关重要。方法 本研究提出了一种基于深度学习的检测系统，可在真实环境中识别不同感染阶段的苹果叶片病害。现有大多数方法侧重于受控条件下的病害检测，而所开发的系统可在光照变化和背景噪声复杂的田间环境中运行。研究开发了两种基于YOLOv8的变体，并集成了多头自注意力（MHSA）模块，以提高对密集、微小和特征相似病斑的检测能力。此外，通过将分割一切模型（SAM）与检测模型相结合，构建了一个基于网络的交互式工具，以支持叶片分离、病害定位、感染估计、严重程度分类和自动报告生成。结果 在所提出的PA-ALeaf数据集上的训练结果表明，与最先进的模型相比，YOLOv8-MHSA_b_h变体取得了最高性能（精确率60.5%，mAP50 56.3%，F1置信度56.0%），同时保持实时推理（每张图像12.38 ms）。结论 通过连接实际病害监测与科学研究，我们的系统为真实果园中的苹果叶片病害检测提供了一种全面、可扩展的解决方案。该系统既可使农民受益，实现多阶段干预，也可为研究人员提供病害进展的见解。","Precision Agriculture",80,{"impact":18,"substance":144,"depth":18,"authority":19,"freshness":105,"relevant":21,"comment":216},"提出面向真实果园的多阶段苹果叶病检测系统并配套网页工具，方法新颖、数据与性能指标明确，对智慧植保具有参考价值。",[218],{"name":213,"url":210},[28,29,30,220,221],"病害识别","苹果病害",[223,224],"苹果叶部病害 深度学习 识别","YOLOv8 苹果病害 检测","苹果叶部病害深度学习识别-2910","10.1007\u002Fs11119-026-10451-5",{"doi":226,"openalex_id":228,"authors":229,"venue":213,"cited_by_count":36,"oa_url":210,"card":240,"direction":47,"ingested_from":50},"W7213537759",[230,232,235,237],{"name":231,"orcid":9},"Kangrui Han",{"name":233,"orcid":234},"Hao Cai","https:\u002F\u002Forcid.org\u002F0000-0002-9879-2960",{"name":236,"orcid":9},"Kari Peter",{"name":238,"orcid":239},"Long He","https:\u002F\u002Forcid.org\u002F0000-0001-9781-6062",{"tldr":241,"method":242,"finding":243,"direction":47,"opportunity":244},"提出基于YOLOv8-MHSA与SAM的网页系统，实现苹果叶片病害多感染阶段实时检测与分级。","YOLOv8+多头自注意力，结合SAM分割，构建PA-ALeaf数据集与网页工具","YOLOv8-MHSA_b_h变体精度60.5%、mAP50 56.3%，单图推理12.38ms，可","可探索轻量化模型在移动端的部署，并融合时序数据预测病害发展轨迹。","2026-09-19T23:30:03.474350Z",{"id":247,"title":248,"url":249,"summary":250,"summary_zh":9,"content":9,"source_name":251,"source_url":9,"published_at":141,"category":12,"cover_url":9,"hotness":63,"is_selected":14,"score":252,"score_detail":253,"sources":255,"tags":257,"search_phrases":260,"slug":263,"view_count":36,"doi":9,"paper":264,"created_at":272},2901,"AgriScope:面向农业图像的像素级多模态理解统一框架,arXiv 2609.20325(预印本)","https:\u002F\u002Farxiv.org\u002Fabs\u002F2609.20325","Boudiaf、Alanssari、Hussain、Javed提出AgriScope,一个统一的像素级多模态农业图像理解框架,联合支持图像级、区域级、像素级理解,可实现接地描述生成、指代表达分割、多轮多模态交互等任务。集成生物专用语义表征、密集空间表征与像素解码;引入大规模像素级农业多模态指令调优数据集AgriGround,包含50万+图像和1100万+指令跟随样本,涵盖植物病害分析、作物与杂草识别、昆虫识别、细粒度植物理解。实验表明AgriScope在多项农业视觉语言任务上有效。","arXiv (preprint)",75,{"impact":18,"substance":144,"depth":18,"authority":105,"freshness":20,"relevant":21,"comment":254},"提出统一像素级农业多模态理解框架并开源50万图像、1100万指令样本的大规模数据集，方法新颖、数据规模突出，但为arXiv预印本、未经同行评审，权威性有限，值得作为前沿技术动态精选。",[256],{"name":251,"url":249},[28,29,258,31,259],"农业遥感","多模态大模型",[261,262],"AgriScope 农业图像 多模态","AgriGround 像素级 农业数据集","AgriScope农业图像多模态-2901",{"doi":9,"openalex_id":9,"authors":265,"venue":9,"cited_by_count":36,"oa_url":9,"card":266,"direction":47,"ingested_from":271},[],{"tldr":267,"method":268,"finding":269,"direction":47,"opportunity":270},"提出AgriScope统一框架，实现农业图像像素级多模态理解与多任务交互。","构建AgriGround数据集（50万+图像、1100万+指令样本），融合语义与","AgriScope在接地描述、指代分割、多轮交互等农业视觉语言任务上有效。","可探索像素级多模态模型在田间实时病害诊断与精准施药决策中的落地与轻量化。","agent","2026-09-19T00:06:08.678379Z"]