[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"topic-SAR":3},{"name":4,"kind":5,"tokens":6,"total":7,"page":8,"page_size":9,"items":10},"SAR","tag",[4],2,1,100,[11,97],{"id":12,"title":13,"url":14,"summary":15,"summary_zh":16,"content":17,"source_name":18,"source_url":14,"published_at":19,"category":20,"cover_url":17,"hotness":21,"is_selected":22,"score":23,"score_detail":24,"sources":30,"tags":32,"primary_topic":36,"secondary_topics":42,"classification":43,"search_phrases":69,"slug":72,"view_count":73,"doi":74,"paper":75,"created_at":96},5571,"Temporal Multi-Scale Spatio-Temporal Network (TMST): Polarization–Temporal Feature Selection and Multi-Scale Spatiotemporal Fusion for Crop Mapping","https:\u002F\u002Fdoi.org\u002F10.3390\u002Fagronomy16202068","Accurate and timely crop mapping supports agricultural resource management, yield prediction, and food security decision-making. In this study, we propose the Temporal Multi-scale Spatio-Temporal Network (TMST), a crop mapping method that integrates multi-source remote sensing time series through polarization–temporal feature selection and multi-scale spatiotemporal fusion. TMST combines Sentinel-1 dual-polarization synthetic aperture radar (SAR), Sentinel-2 multispectral data, and vegetation indices. Its Temporal Feature Importance Explainer (TFIExplainer) selects informative polarization–time features, the Multi-Scale Attention Aggregation (MSAA) module enhances spatial representations, and the Spatio-Temporal Feature Encoder (STFE) models long-range spatial dependencies and phenological dynamics; a composite loss improves minority-class and difficult-pixel recognition. On the Xinxiang dataset, TMST achieved an overall accuracy (OA) of 0.8625 and a mean F1-score (mF1) of 0.8329. Module and data-source ablations suggested benefits of the combined modules and multi-source fusion, while a cropland mask reduced non-cropland misclassification. In direct cross-dataset evaluation, models trained on a self-constructed Austrian dataset and tested on the radar-extended Panoptic Agricultural Satellite Time Series (PASTIS-R) dataset showed that TMST outperformed the evaluated baselines under the specified source–target setting. Annual mapping of Xinxiang for 2019–2023, using year-specific samples and independently trained models, provided map-based estimates of regional cropping structures and spatial patterns. Under the evaluated settings, TMST provides an effective approach to crop classification and annually updated crop mapping.","准确的作物制图有助于农业资源管理、产量预测和粮食安全决策。本研究提出了一种时间多尺度时空网络（Temporal Multi-scale Spatio-Temporal Network，TMST），这是一种通过极化—时间特征选择和多尺度时空融合来整合多源遥感时间序列的作物制图方法。TMST结合了Sentinel-1双极化合成孔径雷达（SAR）、Sentinel-2多光谱数据和植被指数。其时间特征重要性解释器（Temporal Feature Importance Explainer，TFIExplainer）用于选择信息量丰富的极化—时间特征，多尺度注意力聚合（Multi-Scale Attention Aggregation，MSAA）模块增强空间表征，时空特征编码器（Spatio-Temporal Feature Encoder，STFE）建模长程空间依赖关系和物候动态；复合损失函数改善了少数类和困难像元的识别。在新乡数据集上，TMST取得了0.8625的总体精度（OA）和0.8329的平均F1分数（mF1）。模块和数据源消融实验表明，组合模块和多源融合具有优势，而耕地掩膜减少了非耕地误分类。在直接的跨数据集评估中，在自行构建的奥地利数据集上训练并在雷达扩展的Panoptic农业卫星时间序列（PASTIS-R）数据集上测试的模型表明，在指定的源—目标设置下，TMST优于所评估的基线方法。使用逐年样本和独立训练的模型对新乡2019—2023年进行年度制图，提供了基于制图的区域种植结构和空间格局估计。在所评估的设置下，TMST为作物分类和年度更新的作物制图提供了一种有效方法。",null,"Agronomy","2026-10-21T00:00:00Z","论文",10,false,73,{"impact":25,"substance":26,"depth":27,"authority":28,"freshness":21,"relevant":8,"comment":29},12,20,17,14,"提出多源遥感融合的作物制图新方法，方法新颖、实验充分，但属细分技术进展，公共影响有限。",[31],{"name":18,"url":14},[33,34,35,4],"智慧农业","作物制图","遥感",{"code":37,"slug":38,"name":39,"path":40,"parent_code":41},"I3","remote-sensing","遥感与农情监测","\u002Ftopic\u002Fdigital-agriculture\u002Fremote-sensing","I",[],{"status":44,"locked":22,"version":45,"method":46,"evidence":47},"classified","2026-10-22.2","rules",{"candidates":48,"agricultural_evidence":50},[49,56,64],{"code":37,"score":21,"supported":50,"title":51,"abstract":52,"tags":55},true,[],[35,53,54],"remote sensing","multispectral",[35],{"code":57,"score":58,"supported":50,"title":59,"abstract":60,"tags":63},"A8",6,[],[61,62],"粮食安全","food security",[],{"code":65,"score":8,"supported":22,"title":66,"abstract":67,"tags":68},"I1",[],[],[33],[70,71],"Sentinel-1 Sentinel-2 作物分类","TMST 多尺度时空网络","Sentinel-1Sentinel-2作物分类-5571",0,"10.3390\u002Fagronomy16202068",{"doi":74,"openalex_id":76,"authors":77,"venue":18,"cited_by_count":73,"oa_url":14,"card":89,"direction":93,"ingested_from":95},"W7223117203",[78,81,83,86],{"name":79,"orcid":80},"Xi Kan","https:\u002F\u002Forcid.org\u002F0000-0002-4629-9357",{"name":82,"orcid":17},"Wentao Lv",{"name":84,"orcid":85},"Lei Gong","https:\u002F\u002Forcid.org\u002F0000-0001-6429-267X",{"name":87,"orcid":88},"Linglong Zhu","https:\u002F\u002Forcid.org\u002F0000-0002-8863-9911",{"tldr":90,"method":91,"finding":92,"direction":93,"opportunity":94},"提出TMST网络，融合SAR与多光谱时序数据实现作物制图。","Sentinel-1\u002F2时序、极化-时序特征选择、多尺度注意力与时空编码器。","新乡OA达0.8625，多源融合与模块组合有效，跨数据集优于基线。","农业遥感与作物表型","可探索极化-时序特征选择在跨区域迁移与少样本作物制图中的泛化能力。","openalex","2026-10-22T11:20:39.359881Z",{"id":98,"title":99,"url":100,"summary":101,"summary_zh":102,"content":17,"source_name":103,"source_url":100,"published_at":104,"category":20,"cover_url":17,"hotness":21,"is_selected":22,"score":105,"score_detail":106,"sources":110,"tags":112,"primary_topic":116,"secondary_topics":117,"classification":118,"search_phrases":148,"slug":151,"view_count":8,"doi":152,"paper":153,"created_at":191},2786,"SAF-CropNet: Spatially adaptive fusion of SAR and optical imagery for semantic segmentation in operational cropland mapping across regions","https:\u002F\u002Fdoi.org\u002F10.1016\u002Fj.isprsjprs.2026.08.029","Accurate cropland mapping from remote sensing imagery is essential for agricultural monitoring, food security assessment, and land resource management. Although optical and synthetic aperture radar (SAR) observations provide complementary information for cropland extraction, their local contributions vary across imaging conditions, land-cover backgrounds, parcel morphologies, and regional agricultural systems. This variability limits direct concatenation and static fusion, as modality-specific errors may propagate into segmentation outputs in the absence of spatially adaptive weighting. This study proposes SAF-CropNet, a spatially adaptive SAR–optical fusion framework for binary cropland semantic segmentation. Specifically, SAF-CropNet first extracts modality-specific SAR structural features and optical spectral-textural features through separate encoder branches. The Paired Reciprocal Inter-modal Selective Modulator enhances complementary SAR–optical interactions, while the Contextual State-space Modeling module captures long-range parcel organization. A spatially adaptive fusion head then estimates per-pixel contribution weights for SAR, optical, and contextual features before decoding. Evaluation was conducted on a seven-area SAR–optical benchmark spanning China, Germany, and France. The benchmark integrates GF1\u002FGF3 and Sentinel-1\u002F2 imagery with reference labels derived from field surveys and RapidCrops products and covers fragmented smallholder systems, water-rich and peri-urban mosaics, and large mechanized cropland. Within-region five-fold validation shows that SAF-CropNet achieves an average F1 score of 0.8550 and an mIoU of 0.8068, outperforming the evaluated segmentation baselines while remaining lightweight, with 3.38 M parameters and 6.21 GFLOPs for 256 × 256 SAR–optical inputs. In leave-one-area-out transfer experiments, stratified few-shot adaptation increased the average F1 score from 0.4986 under zero-shot transfer to 0.7698. Grouped cross-continental and cross-sensor experiments further showed that direct transfer remains constrained under compound shifts in sensor characteristics, spatial resolution, landscape structure, and reference-label conventions. These findings indicate that SAF-CropNet combines strong within-region segmentation accuracy and computational efficiency with substantial gains from limited target-domain adaptation, while direct zero-shot generalization under severe compound domain shifts remains limited.","从遥感影像中准确提取耕地信息对农业监测、粮食安全评估和土地资源管理至关重要。尽管光学与合成孔径雷达（SAR）观测为耕地提取提供了互补信息，但其局部贡献会随成像条件、地表覆盖背景、地块形态和区域农业系统的不同而变化。这种变异性限制了直接拼接和静态融合的效果，因为在缺乏空间自适应加权的情况下，模态特有的误差可能传播至分割输出中。本研究提出SAF-CropNet，一种用于二分类耕地语义分割的空间自适应SAR–光学融合框架。具体而言，SAF-CropNet首先通过独立的编码器分支提取模态特有的SAR结构特征和光学光谱-纹理特征。成对互反模态间选择性调制器增强SAR与光学之间的互补交互，而上下文状态空间建模模块则捕获长距离地块组织信息。随后，空间自适应融合头在解码前估计SAR、光学和上下文特征的逐像素贡献权重。评估在一个涵盖中国、德国和法国的七区域SAR–光学基准上进行。该基准整合了GF1\u002FGF3和Sentinel-1\u002F2影像，参考标签来源于实地调查和RapidCrops产品，覆盖破碎化小农系统、富水与城郊镶嵌景观以及大规模机械化耕地。区域内五折验证表明，SAF-CropNet平均F1分数达到0.8550，mIoU为0.8068，优于所评估的分割基线，同时保持轻量级，对于256 × 256的SAR–光学输入仅需3.38 M参数和6.21 GFLOPs。在留一区域迁移实验中，分层少样本自适应将平均F1分数从零样本迁移下的0.4986提升至0.7698。分组跨大陆和跨传感器实验进一步表明，在传感器特性、空间分辨率、景观结构和参考标签惯例的复合变化下，直接迁移仍然受到限制。这些发现表明，SAF-CropNet兼具较强的区域内分割精度和计算效率，并能从有限的目标域自适应中获得显著增益，而在严重复合域偏移下的直接零样本泛化能力仍然有限。","ISPRS Journal of Photogrammetry and Remote Sensing","2026-09-17T00:00:00Z",82,{"impact":107,"substance":108,"depth":107,"authority":28,"freshness":21,"relevant":8,"comment":109},18,22,"提出空间自适应SAR-光学融合分割框架，跨七区域基准验证，兼顾精度与轻量化，对耕地遥感制图有实质方法贡献。",[111],{"name":103,"url":100},[113,35,114,115,4],"农业人工智能","耕地监测","作物分类",{"code":37,"slug":38,"name":39,"path":40,"parent_code":41},[],{"status":44,"locked":22,"version":45,"method":46,"evidence":119},{"candidates":120,"agricultural_evidence":50},[121,126,130,137,143],{"code":37,"score":122,"supported":50,"title":123,"abstract":124,"tags":125},8,[],[35,53],[115,35],{"code":57,"score":58,"supported":50,"title":127,"abstract":128,"tags":129},[],[61,62],[],{"code":131,"score":132,"supported":50,"title":133,"abstract":134,"tags":136},"F5",3,[],[135],"机械化",[],{"code":138,"score":132,"supported":50,"title":139,"abstract":140,"tags":142},"G2",[],[141],"smallholder",[],{"code":144,"score":8,"supported":22,"title":145,"abstract":146,"tags":147},"I2",[],[],[113],[149,150],"农业人工智能 作物分类 耕地监测 遥感","农业人工智能 作物分类","农业人工智能作物分类耕地监测遥感-2786","10.1016\u002Fj.isprsjprs.2026.08.029",{"doi":152,"openalex_id":154,"authors":155,"venue":103,"cited_by_count":73,"oa_url":100,"card":186,"direction":93,"ingested_from":95},"W7213463520",[156,159,161,164,167,170,173,175,177,179,181,183],{"name":157,"orcid":158},"minghui chang","https:\u002F\u002Forcid.org\u002F0009-0008-7954-5308",{"name":160,"orcid":17},"Shuaifeng Peng",{"name":162,"orcid":163},"Tao Xu","https:\u002F\u002Forcid.org\u002F0000-0002-7855-4199",{"name":165,"orcid":166},"Yi Yuan","https:\u002F\u002Forcid.org\u002F0009-0006-9481-4938",{"name":168,"orcid":169},"洋一 馬目","https:\u002F\u002Forcid.org\u002F0000-0003-0226-5558",{"name":171,"orcid":172},"Jie Bai","https:\u002F\u002Forcid.org\u002F0000-0003-2426-2358",{"name":174,"orcid":17},"Fugui Luo",{"name":176,"orcid":17},"Jingyu Zhang",{"name":178,"orcid":17},"Xiaoyu Xiao",{"name":180,"orcid":17},"Yu Mu",{"name":182,"orcid":17},"Yong Wang",{"name":184,"orcid":185},"Shihua Li","https:\u002F\u002Forcid.org\u002F0000-0003-4807-5012",{"tldr":187,"method":188,"finding":189,"direction":93,"opportunity":190},"提出SAF-CropNet，用空间自适应融合SAR与光学影像做跨区域耕地语义分割。","双分支编码器提取SAR结构\u002F光学纹理特征，结合互模态调制与状态空间建模，在七区域","区域内F1达0.8550且轻量；零样本迁移F1仅0.4986，少样本适配可升至0.7698。","复合域偏移下零样本泛化差，可研究跨传感器、跨区域的无监督域自适应融合分割方法。","2026-09-17T23:30:34.424579Z"]