IEEE geoscience and remote sensing letters : a publication of the IEEE Geoscience and Remote Sensing Society最新文献_第5页

Spectral Descriptor-Based Retrieval of Localized Events for Infrasound Time Series Signals 基于谱描述子的次声时间序列局部事件检索

IF 4.4

IEEE geoscience and remote sensing letters : a publication of the IEEE Geoscience and Remote Sensing Society Pub Date : 2026-01-01 Epub Date: 2026-08-06 DOI: 10.1109/LGRS.2026.3720906

Iustin Floroiu;Andrei Anghel;Mihai Datcu;Laura Petrescu;Cosmin Dănişor

{"title":"Spectral Descriptor-Based Retrieval of Localized Events for Infrasound Time Series Signals","authors":"Iustin Floroiu;Andrei Anghel;Mihai Datcu;Laura Petrescu;Cosmin Dănişor","doi":"10.1109/LGRS.2026.3720906","DOIUrl":"https://doi.org/10.1109/LGRS.2026.3720906","url":null,"abstract":"Atmospheric infrasound archives contain short, impulsive events, sustained machinery, microbarom regimes, storms, and front-driven turbulence whose relevant time scales range from seconds to days. Representing all such phenomena as fixed-length, single-label examples can hide the internal spectral evolution that often defines long-duration atmospheric processes. This letter formulates localized infrasound discovery as a query-by-example retrieval in a short-time spectral descriptor space. Each local window is represented by five descriptors: spectral centroid, flux, rolloff, entropy, and zero-crossing count. Longer query and candidate intervals are encoded as ordered descriptor sequences, standardized using archive-estimated feature statistics, and ranked by aligned Euclidean sequence distance. The method is validated on synthetic regimes and with known ground truth, and compared against a mel-frequency cepstral coefficient (MFCC) baseline under the same retrieval protocol. Precision metrics were used to assess the retrieval performance. Real coastal infrasound examples from Agigea, Constanţa, Romania, show that the same representation retrieves storm-consistent spectral morphology across temporally separated intervals while rejecting dissimilar high-energy variability.","PeriodicalId":91017,"journal":{"name":"IEEE geoscience and remote sensing letters : a publication of the IEEE Geoscience and Remote Sensing Society","volume":"23 ","pages":"1002305-1002305"},"PeriodicalIF":4.4,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11643806","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148836669","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Boosting Large Vision-Language Models With Hierarchical Semantic Fusion for Remote Sensing Visual Question Answering 基于层次语义融合的大型视觉语言模型在遥感视觉问答中的应用

IF 4.4

IEEE geoscience and remote sensing letters : a publication of the IEEE Geoscience and Remote Sensing Society Pub Date : 2026-01-01 Epub Date: 2026-03-07 DOI: 10.1109/LGRS.2026.3691060

Fan Liu;Jian Huo;Dongmei Wang;Yihong Wang;Yi Xu;Wenwen Dai;Yuexuan An

{"title":"Boosting Large Vision-Language Models With Hierarchical Semantic Fusion for Remote Sensing Visual Question Answering","authors":"Fan Liu;Jian Huo;Dongmei Wang;Yihong Wang;Yi Xu;Wenwen Dai;Yuexuan An","doi":"10.1109/LGRS.2026.3691060","DOIUrl":"https://doi.org/10.1109/LGRS.2026.3691060","url":null,"abstract":"Remote sensing (RS) visual question answering (VQA) aims to answer natural language queries by jointly modeling visual content and semantic meaning in RS imagery. Large vision-language models (LVLMs) offer strong visual understanding and reasoning capabilities and provide a promising foundation for VQA tasks. However, existing LVLMs are not well adapted to RS scenes, as deep-layer focused visual representations tend to overlook low- and mid-level structural cues required for RS-specific reasoning. To address this limitation, we propose the hierarchical semantic fusion network, hierarchical semantic fusion network (HSF-Net), a remote-sensing-oriented enhancement framework built upon LLaVA-v1.5 to balance the detail perception and semantic reasoning in RS-VQA. HSF-Net integrates three complementary modules: 1) frequency-selective unsharp masking (FS-USM) amplifies high-frequency structural details in RS imagery; 2) hierarchical cross-level semantic fusion (HCSF) injects multiscale shallow and mid-level cues into deep semantic features while preserving language alignment; 3) self-attention-based semantic reweighting (SRS) refines internal semantic correlations to strengthen contextual understanding. Extensive experimental evaluations on three benchmark RS-VQA datasets, compared with lightweight specialized models and superior LVLMs, demonstrate HSF-Net’s state-of-the-art (SOTA) performance across all tasks. Our code is available at: <uri>https://github.com/HUOUO7/HSF-Net</uri>","PeriodicalId":91017,"journal":{"name":"IEEE geoscience and remote sensing letters : a publication of the IEEE Geoscience and Remote Sensing Society","volume":"23 ","pages":"6010505-6010505"},"PeriodicalIF":4.4,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148484958","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Proximal Sensing With Rotating Panoramic Field Cameras: An Open-Source Platform for Remote Sensing Validation 近端遥感与旋转全景场相机：遥感验证的开源平台

IF 4.4

IEEE geoscience and remote sensing letters : a publication of the IEEE Geoscience and Remote Sensing Society Pub Date : 2026-01-01 Epub Date: 2026-03-11 DOI: 10.1109/LGRS.2026.3673199

Florian J. Ellsäßer;Claudia Paris;Michael Schlund;Harald van der Werff;Anton Vrieling

{"title":"Proximal Sensing With Rotating Panoramic Field Cameras: An Open-Source Platform for Remote Sensing Validation","authors":"Florian J. Ellsäßer;Claudia Paris;Michael Schlund;Harald van der Werff;Anton Vrieling","doi":"10.1109/LGRS.2026.3673199","DOIUrl":"https://doi.org/10.1109/LGRS.2026.3673199","url":null,"abstract":"Proximal sensing systems provide high-resolution, continuous observations that are essential for environmental monitoring and for validating space- and airborne remote sensing products. However, conventional field cameras typically have limited fields of view, which reduces their spatial representativeness. To address this limitation, we developed VegTrack360, an open-source, modular two-axis rotating camera prototype system that generates full 360° panoramic imagery from a single narrow-field-of-view sensor. VegTrack360 was field-tested using an RGB camera to validate the panoramic acquisition concept at agricultural and mixed-use sites in the Netherlands. Although the system is designed for flexible integration of interchangeable RGB, thermal infrared, or multispectral sensors, this study focuses on an RGB-based proof-of-concept evaluation. VegTrack360 was benchmarked against a commercial panoramic camera, a stationary PhenoCam, and drone/uncrewed aerial system (UAS)-based RGB orthomosaics. The results demonstrate high color fidelity (histogram correlation r <inline-formula> <tex-math>$ge 0.99$ </tex-math></inline-formula>) and good radiometric stability. Greenness time series derived from two VegTrack360 observation positions showed strong agreement with the co-located PhenoCam (r = 0.92 and r = 0.72), while differences with UAS products mainly reflected contrasting viewing geometries. Overall, VegTrack360 complements existing proximal and airborne sensing approaches by providing an open-source, sensor-interchangeable platform for 360° panoramic vegetation monitoring.","PeriodicalId":91017,"journal":{"name":"IEEE geoscience and remote sensing letters : a publication of the IEEE Geoscience and Remote Sensing Society","volume":"23 ","pages":"1-5"},"PeriodicalIF":4.4,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11430529","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148486188","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Directional Object Detection in Aerial Images 航拍图像中的定向目标检测

IF 4.4

IEEE geoscience and remote sensing letters : a publication of the IEEE Geoscience and Remote Sensing Society Pub Date : 2026-01-01 Epub Date: 2026-03-14 DOI: 10.1109/LGRS.2026.3693472

Szilard Molnar;Levente Tamas;Zoltan Kato

引用次数: 0

FALL: Fine-Grained Alignment and Local Information Aggregation Learning for Drone-Satellite Cross-View Geo-Localization 无人机-卫星交叉视点地理定位的细粒度对齐和局部信息聚合学习

IF 4.4

IEEE geoscience and remote sensing letters : a publication of the IEEE Geoscience and Remote Sensing Society Pub Date : 2026-01-01 Epub Date: 2026-03-07 DOI: 10.1109/LGRS.2026.3691132

Haolin Du;Jingfei He;Zihan Nan;Chenghu Mi;Zihao Chen

引用次数: 0

GP-BO-Driven Ensemble Learning for High-Resolution Surface Soil Moisture Retrieval gp - bo驱动的高分辨率表层土壤水分反演集成学习

IF 4.4

IEEE geoscience and remote sensing letters : a publication of the IEEE Geoscience and Remote Sensing Society Pub Date : 2026-01-01 Epub Date: 2025-12-02 DOI: 10.1109/LGRS.2025.3639405

Jianhui Chen;Zuo Wang;Ping Wei;Chang Huang;Lisheng Song;Yuanhong You;Shuoqi Zhang;Zhijie Dong

{"title":"GP-BO-Driven Ensemble Learning for High-Resolution Surface Soil Moisture Retrieval","authors":"Jianhui Chen;Zuo Wang;Ping Wei;Chang Huang;Lisheng Song;Yuanhong You;Shuoqi Zhang;Zhijie Dong","doi":"10.1109/LGRS.2025.3639405","DOIUrl":"https://doi.org/10.1109/LGRS.2025.3639405","url":null,"abstract":"Surface soil moisture (SSM) plays a crucial role in hydrological processes, ecosystem dynamics, and agricultural management. Currently, high-spatial-resolution SSM estimation primarily relies on machine learning methods. However, in heterogeneous environments, the challenges associated with hyperparameter optimization, computational efficiency, and uncertainty control compromise the robustness of these methods. To address this issue, this study introduces and evaluates an integrated strategy that combines Gaussian process Bayesian optimization (GP-BO) with machine learning for high-resolution SSM retrieval. The results demonstrate that the combined method significantly outperforms conventional optimization methods evaluated by the test sets from Heihe River Basin, Naqu, and Shandian River basins. Notably, the integration of GP-BO with extreme gradient boosting (XGBoost) turned out to be the optimal combination, improving <inline-formula> <tex-math>$R^{2}$ </tex-math></inline-formula> by 0.01–0.19 and reducing unbiased root-mean-square error (ubRMSE) by 0.02–1.71 percentage points relative to conventional optimizers, while requiring the least training time for ensemble models. Furthermore, vegetation-specific GP-BO-tuned XGBoost models achieve varying degrees of accuracy improvement and reduced uncertainty across various vegetation, particularly in barren and grassland regions. These findings highlight the effectiveness of the GP-BO hyperparameter optimization algorithm in reducing the uncertainties of SSM estimation in heterogeneous environments.","PeriodicalId":91017,"journal":{"name":"IEEE geoscience and remote sensing letters : a publication of the IEEE Geoscience and Remote Sensing Society","volume":"23 ","pages":"1-5"},"PeriodicalIF":4.4,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148486274","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

EAR-Net: Edge-Guided Adapter-Enhanced Recursive Fusion Network for Small-Object Detection in Remote Sensing Imagery EAR-Net：用于遥感图像小目标检测的边缘引导适配器增强递归融合网络

IF 4.4

IEEE geoscience and remote sensing letters : a publication of the IEEE Geoscience and Remote Sensing Society Pub Date : 2026-01-01 Epub Date: 2026-03-15 DOI: 10.1109/LGRS.2026.3693693

Yan Zhang;Qiang Wang;Ruihan Bai;Hui Li

{"title":"EAR-Net: Edge-Guided Adapter-Enhanced Recursive Fusion Network for Small-Object Detection in Remote Sensing Imagery","authors":"Yan Zhang;Qiang Wang;Ruihan Bai;Hui Li","doi":"10.1109/LGRS.2026.3693693","DOIUrl":"https://doi.org/10.1109/LGRS.2026.3693693","url":null,"abstract":"Small-object detection in remote sensing imagery is challenging because tiny targets often suffer from large-scale variation, weak boundaries, and heavy background interference. Although modern one-stage detectors are efficient, repeated downsampling often removes fine boundary details. As a result, dense tiny objects may become difficult to separate in deep feature maps. Meanwhile, background textures such as roads, roofs, and shadows may look similar to small targets, leading to false detections. To address these problems, we propose EAR-Net, an edge-guided adapter-enhanced recursive fusion network built on YOLOv11n. EAR-Net improves small-object representation from three aspects. First, the multiscale edge-guided (MSEG) feature injection module uses Sobel-initialized learnable edge operators to provide explicit boundary guidance. Second, the adapter-enhanced C2PSA module introduces lightweight bottleneck adapters into the attention branch to suppress high-frequency background noise while preserving global context. Third, the recursive feature fusion mechanism introduces a recursive “focus–distribute–refocus–fuse” process to reduce cross-scale spatial misalignment. Experiments on NWPU VHR-10, DIOR, RSOD, and HRRSD show that EAR-Net improves detection accuracy and localization precision while maintaining a practical balance between accuracy and efficiency.","PeriodicalId":91017,"journal":{"name":"IEEE geoscience and remote sensing letters : a publication of the IEEE Geoscience and Remote Sensing Society","volume":"23 ","pages":"6012205-6012205"},"PeriodicalIF":4.4,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148486309","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Knowledge-Enhanced RT-DETR for Infrared Small Target Detection Driven by Multilevel Domain Priors 基于多阶域先验的知识增强RT-DETR红外小目标检测

IF 4.4

IEEE geoscience and remote sensing letters : a publication of the IEEE Geoscience and Remote Sensing Society Pub Date : 2026-01-01 Epub Date: 2026-06-29 DOI: 10.1109/LGRS.2026.3708198

Xiang Ji;Gang Liu;Zhanyang Gao;Jungang Yan;Shuning Cao

{"title":"Knowledge-Enhanced RT-DETR for Infrared Small Target Detection Driven by Multilevel Domain Priors","authors":"Xiang Ji;Gang Liu;Zhanyang Gao;Jungang Yan;Shuning Cao","doi":"10.1109/LGRS.2026.3708198","DOIUrl":"https://doi.org/10.1109/LGRS.2026.3708198","url":null,"abstract":"Infrared small target detection (IRSTD) remains an arduous task characterized by diminished target signatures, suppressed signal-to-noise ratios, and intricate environmental clutter. Conventional data-driven paradigms often fail to resolve dim targets from overwhelming background interference, as the weak signals are easily submerged by noise. In this study, we present a knowledge-enhanced RT-DETR (KE-RT-DETR) framework that synergistically integrates infrared physical priors into a multilevel neural architecture, facilitating robust target-background separation in low-SNR scenarios. At the input stage, a statistical-physical context enhancement (SPCE) module is employed to augment target saliency by capturing directional responses and statistical distributions. This is complemented by a soft-gated SCConv (SG-SCConv) during feature extraction, which effectively decouples high-frequency target components from low-frequency background fluctuations to strengthen feature discriminability. Furthermore, a gradient-adaptive reparameterized central difference convolution (GA-RepCDC) is introduced in the feature fusion layer to accentuate local contrast through differential modeling while maintaining high inference throughput via structural reparameterization. The experimental results on the DAUB and IRSTD-1k datasets demonstrate that the proposed method achieves superior performance against the compared methods. The code and dataset will be made publicly available at <uri>https://github.com/Jixiang975231/KE-RT-DETR</uri>","PeriodicalId":91017,"journal":{"name":"IEEE geoscience and remote sensing letters : a publication of the IEEE Geoscience and Remote Sensing Society","volume":"23 ","pages":"7001605-7001605"},"PeriodicalIF":4.4,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148486472","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

AnisoColSAR: A Physics-Informed Approach for SAR Image Colorization by Encoding Scattering Anisotropy AnisoColSAR：基于散射各向异性编码的SAR图像着色方法

IF 4.4

IEEE geoscience and remote sensing letters : a publication of the IEEE Geoscience and Remote Sensing Society Pub Date : 2026-01-01 Epub Date: 2026-03-31 DOI: 10.1109/LGRS.2026.3679380

Xinwei An;Wei Yang;Anqi Gao;Hongcheng Zeng;Wanting Yang;Yamin Wang;Jie Chen

{"title":"AnisoColSAR: A Physics-Informed Approach for SAR Image Colorization by Encoding Scattering Anisotropy","authors":"Xinwei An;Wei Yang;Anqi Gao;Hongcheng Zeng;Wanting Yang;Yamin Wang;Jie Chen","doi":"10.1109/LGRS.2026.3679380","DOIUrl":"https://doi.org/10.1109/LGRS.2026.3679380","url":null,"abstract":"Synthetic aperture radar (SAR) products are mostly presented as grayscale images, where the scattering characteristics of ground objects lack intuitive visualization, limiting the performance of downstream interpretation tasks. To address this issue, this letter proposes AnisoColSAR, a physics-informed SAR image colorization method. Based on the physical scattering mechanism of ground objects, AnisoColSAR takes the scattering anisotropy index (SAI) as a proxy for physical anisotropy. By quantifying and encoding the SAI into color features, it generates pseudo-color images with explicit physical significance without full-polarization or multisource auxiliary data. Experiments on the AIR-PolSAR-Seg-2.0 dataset demonstrate that the proposed pseudo-color images stably improve semantic segmentation performance across all polarization modes (HH, VV, HV, VH), with particularly significant gains in cross-polarization modes. This study provides a novel physics-guided approach for SAR image interpretation, and the generated pseudo-color images are promising for various remote sensing tasks such as classification and detection, showing strong practical value.","PeriodicalId":91017,"journal":{"name":"IEEE geoscience and remote sensing letters : a publication of the IEEE Geoscience and Remote Sensing Society","volume":"23 ","pages":"1-5"},"PeriodicalIF":4.4,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148486473","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A Deep U-Net Kriging Framework for Interpolating Remote Sensing Images 一种用于遥感图像插值的深度U-Net Kriging框架

IF 4.4

IEEE geoscience and remote sensing letters : a publication of the IEEE Geoscience and Remote Sensing Society Pub Date : 2026-01-01 Epub Date: 2026-06-29 DOI: 10.1109/LGRS.2026.3708253

João Pedro Malim Franco;Vitor Salomão Recacho;Guilherme Colombo Soares;Márcio Poletti Laurini

{"title":"A Deep U-Net Kriging Framework for Interpolating Remote Sensing Images","authors":"João Pedro Malim Franco;Vitor Salomão Recacho;Guilherme Colombo Soares;Márcio Poletti Laurini","doi":"10.1109/LGRS.2026.3708253","DOIUrl":"https://doi.org/10.1109/LGRS.2026.3708253","url":null,"abstract":"Spatial interpolation is a fundamental task in remote sensing, allowing the reconstruction of continuous geophysical fields from sparsely sampled observations. Kriging has long been considered the gold standard, owing to its optimal linear prediction properties under second-order stationarity. Nevertheless, it often struggles to capture nonlinear residual structures and fine-scale heterogeneities. In this letter, we introduce a two-stage hybrid framework that integrates ordinary kriging (OK) for baseline interpolation with a deep U-Net convolutional neural network (CNN) applied to the residuals. The kriging component accounts for large-scale spatial dependence through a variogram-based predictor, while the U-Net enhances the interpolation by learning high-frequency, nonlinear corrections. Experiments on remote sensing imagery demonstrate that the proposed hybrid approach substantially improves upon classical kriging and achieves the best mean absolute error (MAE) among the evaluated methods, while remaining competitive in terms of mean squared error (mse).","PeriodicalId":91017,"journal":{"name":"IEEE geoscience and remote sensing letters : a publication of the IEEE Geoscience and Remote Sensing Society","volume":"23 ","pages":"8001905-8001905"},"PeriodicalIF":4.4,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148486514","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0