IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing最新文献

{"title":"Coseismic and Postseismic Slip Analysis of the 2024 Mw 7.0 Wushi Earthquake: Insights From InSAR Observations and Slip Inversion","authors":"Jianlong Chen;Zicheng Huang;Lejun Lu;Peizhen Zhang","doi":"10.1109/JSTARS.2025.3602843","DOIUrl":"https://doi.org/10.1109/JSTARS.2025.3602843","url":null,"abstract":"Analysis of fault slip behavior is crucial for assessing regional seismic hazards. This study employs Sentinel-1 SAR data to investigate the coseismic and postseismic deformation associated with the 2024 <italic>M</i>w 7.0 Wushi earthquake using Interferometric Synthetic Aperture Radar (InSAR) technique. Through multitemporal InSAR analysis, we reconstructed the coseismic slip distribution and revealed the spatiotemporal evolution of postseismic slip. The coseismic rupture released a seismic moment of 3.65<inline-formula><tex-math>$times 10^{19}$</tex-math></inline-formula> N <inline-formula><tex-math>$cdot$</tex-math></inline-formula> m (<italic>M</i>w 7.0) with maximum slip of 2.3 m at 12 km depth. Coseismic Coulomb stress changes triggered the <italic>M</i> 5.6 aftershock on January 30, 2024, with surface rupture observed. Postseismic deformation exhibits coupled afterslip–aftershock mechanisms, showing asymmetric line of sight (LOS) displacement with maximum uplift of <inline-formula><tex-math>$sim$</tex-math></inline-formula>20 cm (ascending track) and <inline-formula><tex-math>$sim$</tex-math></inline-formula>15 cm subsidence superimposed on 20 cm uplift (descending track). Time-series analysis reveals a two-phase decay pattern: rapid deformation at 20 cm/month during the initial two weeks, decaying to 0.5 cm/month over subsequent four months. The friction parameter (0.001 <inline-formula><tex-math>$&lt; $</tex-math></inline-formula> <inline-formula><tex-math>$a - b$</tex-math></inline-formula> <inline-formula><tex-math>$&lt; $</tex-math></inline-formula> 0.003) indicates velocity-strengthening behavior in the Maidan fault’s postseismic slip zone, suggesting stable slip characteristics. However, the relatively low friction coefficient still warrants attention to long-term fault motion and strain accumulation. This study provides new observational constraints for understanding postseismic deformation mechanisms and frictional properties in the South Tianshan orogenic belt.","PeriodicalId":13116,"journal":{"name":"IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing","volume":"18 ","pages":"22620-22629"},"PeriodicalIF":5.3,"publicationDate":"2025-08-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11141694","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145089965","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Early Detection of Forest Disturbances in Homogeneous Stands – Deep Learning-Based Monitoring of Picea Abies","authors":"Maximilian Kirsch;Jakob Wernicke;Pawanjeet Singh Datta;Christine Preisach","doi":"10.1109/JSTARS.2025.3603094","DOIUrl":"https://doi.org/10.1109/JSTARS.2025.3603094","url":null,"abstract":"Climate change has increased the vulnerability of forests to insect-related damage, resulting in widespread forest loss in Central Europe and highlighting the need for effective, continuous monitoring systems. Remote sensing-based approaches, often rely on supervised machine learning (ML) algorithms that require labeled training data. Monitoring temporal patterns through time series analysis offers a potential alternative for earlier detection of disturbances but requires substantial storage resources. This study investigates the potential of a Deep Learning algorithm based on a Long Short Term Memory (LSTM) Autoencoder for the detection of anomalies in forest health (e.g., bark beetle outbreaks), utilizing Sentinel-2 time series data. This approach is an alternative to supervised ML methods, avoiding the necessity for labeled training data. Furthermore, it is more memory-efficient than other time series analysis approaches, as a robust model can be created using only a 26-week-long time series as input. In this study, we monitored pure stands of spruce in Thuringia, Germany, over a 7-year period from 2018 to the end of 2024. Our best model achieved a detection accuracy of 87% on test data and was able to detect 65% of all anomalies at a very early stage (more than a month before visible signs of forest disturbance). Compared to the established time series break detection algorithm – breaks for additive season and trend and isolation forest anomaly detection, our approach consistently detected a higher percentage of anomalies at an earlier stage. These findings suggest that LSTM-based Autoencoders could provide a promising, resource-efficient approach to forest health monitoring, enabling more timely responses to emerging threats.","PeriodicalId":13116,"journal":{"name":"IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing","volume":"18 ","pages":"22300-22316"},"PeriodicalIF":5.3,"publicationDate":"2025-08-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11141690","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145089977","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Adaptive Multiorder Graph Regularized NMF With Dual Sparsity for Hyperspectral Unmixing","authors":"Hui Chen;Liangyu Liu;Xianchao Xiu;Wanquan Liu","doi":"10.1109/JSTARS.2025.3602505","DOIUrl":"https://doi.org/10.1109/JSTARS.2025.3602505","url":null,"abstract":"Hyperspectral unmixing (HU) is a critical yet challenging task in remote sensing. However, existing nonnegative matrix factorization (NMF) methods with graph learning mostly focus on first-order or second-order nearest neighbor relationships and usually require manual parameter tuning, which fails to characterize intrinsic data structures. To address the above issues, we propose a novel adaptive multiorder graph regularized NMF method (MOGNMF) with three key features. First, multiorder graph regularization is introduced into the NMF framework to exploit global and local information comprehensively. Second, these parameters associated with the multiorder graph are learned adaptively through a data-driven approach. Third, dual sparsity is embedded to obtain better robustness, i.e., <inline-formula><tex-math>$ell _{1/2}$</tex-math></inline-formula>-norm on the abundance matrix and <inline-formula><tex-math>$ell _{2,1}$</tex-math></inline-formula>-norm on the noise matrix. To solve the proposed model, we develop an alternating minimization algorithm whose subproblems have explicit solutions, thus ensuring effectiveness. Experiments on simulated and real hyperspectral data indicate that the proposed method delivers better unmixing results.","PeriodicalId":13116,"journal":{"name":"IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing","volume":"18 ","pages":"22121-22136"},"PeriodicalIF":5.3,"publicationDate":"2025-08-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11141691","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145060959","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"MRC-Net: A Multistage Network With Rank-Reduced Multihead Self-Attention and Cascade Learning for Hyperspectral Pansharpening","authors":"Chanyue Wu;Rui Feng;Pei Zhang;Hanyu Mao;Dong Wang;Zongwen Bai;Ying Li","doi":"10.1109/JSTARS.2025.3603113","DOIUrl":"https://doi.org/10.1109/JSTARS.2025.3603113","url":null,"abstract":"Hyperspectral (HS) pansharpening aims to enhance the spatial structure of Low-Resolution (LR) HS images guided by High-Resolution (HR) Panchromatic (PAN) images while maintaining spectral fidelity of resulting HR–HS images. Even though single-stage architecture dominates current state-of-the-art approaches and delivers competitive performance, they may inadequately capture the intricate spatio-spectral dependencies in HR–HS images in only a single stage. In addition, the redundant information of HS images incurs heavy computational overhead. To this end, we propose a Multistage Network with Rank-reduced multihead self-attention and Cascade learning (MRC-Net), which progressively reconstructs the ideal HR–HS images. Our method gradually incorporates high-frequency details from the PAN image into the LR–HS images through multiple stages, enabling the comprehensive learning and processing of the interactive relationship between high-dimensional spectral and complex spatial structures. In addition, we introduce a rank-reduced multihead self-attention module at substage levels. As such, redundant information of HS images can be reduced so as to alleviate computational burden, and the learning capacity of MRC-Net is improved to capture long-range dependencies and enhance global contextual relevance. To further obtain the high-dimensional spatio-spectral relationship, we optimize the MRC-Net using a cascade loss function with adaptive weighting across all stages, which guides the learning process and improves the overall performance. Extensive evaluations on three publicly available datasets (Pavia Centre, Chikusei, and Houston 2018) demonstrate that the proposed MRC-Net outperforms the state-of-the-art pansharpening methods in both quantitative comparison and qualitative analysis, which highlights its effectiveness in both spatial enhancement and spectral preservation.","PeriodicalId":13116,"journal":{"name":"IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing","volume":"18 ","pages":"22466-22485"},"PeriodicalIF":5.3,"publicationDate":"2025-08-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11141687","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145090013","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Semiautomatic Workflow for Accurate LiDAR-Derived DSM Retrieval in Aquatic Scenarios via Water Surface Mapping","authors":"Erika Piaser;Andrea Berton;Giovanna Sona;Paolo Villa","doi":"10.1109/JSTARS.2025.3602156","DOIUrl":"https://doi.org/10.1109/JSTARS.2025.3602156","url":null,"abstract":"High-resolution, 3-D water surface mapping in aquatic environments is critical for evaluating complex interactions between human activities and environmental dynamics. Despite the overall potential of LiDAR data to generate 3-D point clouds, providing the accurate and complete digital surface models (DSMs) at the interface between terrestrial and aquatic ecosystems is still a significantly challenging task. In fact, due to water’s strong near-infrared absorption and its near specularity, LiDAR often results in weak or missing signal returns. In addition, direct linear interpolation for gap filling can introduce biases in the DSM reconstruction, especially near shorelines. This study proposes a four-step semiautomatic, open-source workflow for high-resolution DSM reconstruction in aquatic scenarios, using unsupervised machine learning for land–water classification based on optimized LiDAR-derived features. Mean water-level surface elevation, extracted from binary scene clustering, was used to fill DSM gaps over water via ad hoc gap filling. The accuracy of the resulting “water-filled” DSM (WFDSM) was evaluated across six diverse real-world aquatic scenarios with a range of challenging conditions (e.g., presence of aquatic vegetation, detached ponds, man-made structures, and land depressions) and compared against open-source products. Unsupervised clustering combining radiometric and geometric features achieved high classification accuracy (<italic>F</i>-score > 0.97) for “water-level” targets, with negligible commission errors. Unlike standard products, WFDSM effectively handles variations in terrain and surface, maintaining low elevation biases (< 25 cm) even in areas with complex vegetation and fine-scale anthropogenic structures, thus demonstrating high suitability in both transitional and open-water areas.","PeriodicalId":13116,"journal":{"name":"IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing","volume":"18 ","pages":"22673-22689"},"PeriodicalIF":5.3,"publicationDate":"2025-08-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11141703","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145051016","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Optimization Method for Remote Sensing Image-Based Photovoltaic Panel Segmentation via Perception-Driven Enhancement in Nonideal Environments","authors":"Xuewei Chao;Lixin Zhang;Yang Li;Jing Nie;Shuo Yang;Sezai Ercisli","doi":"10.1109/JSTARS.2025.3602477","DOIUrl":"https://doi.org/10.1109/JSTARS.2025.3602477","url":null,"abstract":"This research addresses the challenges of feature distortion, poor scale adaptability, and high model complexity in the segmentation of photovoltaic panels from remote sensing images under nonideal conditions, such as rain and fog. To tackle these issues, a DMFA-DeepLab model integrating perception-driven enhancement is proposed. First, a physical perception degradation data generation method is developed based on the CycleGAN framework to simulate optical degradation features caused by rain and fog, thereby enhancing the model’s generalization capability in adverse environmental conditions. Second, a multiscale convolutional attention module (MCAM) is designed, which captures cross-scale features through heterogeneous convolution branches with receptive fields of 3, 5, and 9. The module further incorporates a channel–spatial dual-attention mechanism to dynamically focus on key regions while suppressing background interference. Based on MCAM, a multilevel feature aggregation network is constructed, which enhances boundary description through cross-level feature fusion. To achieve model lightweighting, a two-stage pruning–knowledge distillation strategy is introduced: initially, 30% of low-contribution channels are pruned based on the BN scaling factor through sparse training; after a second pruning of 20%, the cumulative compression rate reaches 44% . Finally, knowledge distillation is applied, where the original model serves as the teacher to guide the student model in performance restoration. The experimental results demonstrate that, on the enhanced dataset simulating rain and fog environments, the complete model achieves an MIoU of 93.17%, representing an improvement of 7.04% over the baseline DeepLabV3+. The lightweight model DMFA-DeepLab, while reducing the parameter count by 44%, restores the MIoU to 92.96% and increases inference speed by 2.3 times. Compared with the mainstream models, such as U-Net and PSPNet, DMFA-DeepLab demonstrates significantly superior segmentation accuracy and robustness in complex environments, achieving an <italic>F</i>1-score of 94.57%, which is 4.67% points higher than the second-best model.","PeriodicalId":13116,"journal":{"name":"IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing","volume":"18 ","pages":"22513-22529"},"PeriodicalIF":5.3,"publicationDate":"2025-08-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11139120","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145090214","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A Spaceborne Radar-Optical Joint Localization and Measurement Scheme for Enhancing Noncooperative Spacecraft Perception","authors":"Xindi Yu;Ling Wang;Bin Wu;Lutong Zhou;Daiyin Zhu","doi":"10.1109/JSTARS.2025.3602155","DOIUrl":"https://doi.org/10.1109/JSTARS.2025.3602155","url":null,"abstract":"Localization and measurement of noncooperative spacecrafts in space are of great significance for on-orbit flight safety, rendezvous, and other operations. The multisensor fusion-based perception has gained much interest in existing space situational awareness systems, among which the radar-optical fusion is very typical. The all-day persistent detection and perception ability of the radar is complemented to the regular optical image perception. However, few methods have been explicitly reported on how to deeply fuse the radar and optical data to enhance the perception capability. This article presents a spaceborne radar-optical joint perception scheme to enhance the situation awareness ability of the on-orbit flying spacecraft, focusing on the accurate far-range localization of the approaching spacecraft and measurement of its key components by leveraging the orthogonal observation geometry provided by the radar and camera imaging. A “one-radar, three-camera” configuration is adopted to support single-platform, coplatform and cross-platform sensing with geometric flexibility. The localization stage combines radar slant range with optical azimuth and elevation angles through a spatiotemporal synchronization. Under a 5 dB signal-to-noise ratio, the method achieves an average relative localization error of 0.3%. For component measurement, three sensing modes are designed and fused using Dempster-Shafer theory to address illumination and range challenges: 1) radar-only, 2) coplatform radar-optical, and 3) cross-platform radar-optical. Experimental results show that under ideal conditions, the proposed system achieves length and orientation estimation errors within 3% and 2°, respectively. Even in poor lighting or far-range scenarios, radar ensures measurement reliability and improves accuracy by over 10% compared to strategies involving optical sensing. Although optical-only experiments are not performed, the proposed framework is compatible with optical-only inputs due to its modular structure.","PeriodicalId":13116,"journal":{"name":"IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing","volume":"18 ","pages":"21855-21868"},"PeriodicalIF":5.3,"publicationDate":"2025-08-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11134817","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145061867","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Automated Extraction of Impervious Surface Area Using Hyperlocal Samples From Multisource Data Fusion Across Economic–Geographic Zones","authors":"Nan Wang;Yunlan Guan;Yuqian Wang;Qirui Fang;Zixuan Li;Jian Dong;Jing Luo","doi":"10.1109/JSTARS.2025.3602036","DOIUrl":"https://doi.org/10.1109/JSTARS.2025.3602036","url":null,"abstract":"Due to their complex spatial structure and spectral heterogeneity, mapping impervious surface area (ISA) at large scales with high resolution remains challenging. Key issues include labor-intensive sample collection, suboptimal feature selection that often fails to account for regional climate and urban development differences, and the lack of up-to-date ISA products at 10 m resolution. To address these limitations, we propose an automated ISA extraction method based on economic–geographic zoning and multisource data fusion using hyperlocal samples. The study area is first divided into zones based on climate zones and urban development levels. Within each zone, hexagonal units are defined to enable localized sample selection and feature optimization. Training samples are automatically generated using rule-based and threshold-based methods, incorporating land use, land cover, and human activity information from OpenStreetMap, Sentinel-2 imagery, and Black Marble nighttime light (NTL) imagery. A total of 45 features are constructed, with NTL features playing an innovative role in improving classification performance. The recursive feature elimination with cross-validation algorithm is applied to select the most relevant features for each zone. Based on the Google Earth engine platform, a regionally adaptive random forest model is implemented to produce a 10-m-resolution ISA map of the study area. The results show that the proposed hyperlocal sampling strategy significantly improves sample quality. Zonal feature selection not only simplifies the model but also enhances interpretability and classification accuracy. Quantitative analysis confirms that NTL features are among the most significant inputs. The proposed method achieves an overall accuracy of 90.72%, outperforming existing ISA products GAIA, FROM_GLC10, GHSL, and ESRI_LandCover by 7.12%, 3.02%, 5.07%, and 2.47%, respectively, demonstrating its effectiveness.","PeriodicalId":13116,"journal":{"name":"IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing","volume":"18 ","pages":"22602-22619"},"PeriodicalIF":5.3,"publicationDate":"2025-08-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11134770","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145049811","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"LAI Inversion for Winter Wheat From UAV Hyperspectral Data Considering Crop Growth Heterogeneity","authors":"Jiahui Feng;Zhaozhao Zeng;Yuyun Liang;Jun Li","doi":"10.1109/JSTARS.2025.3602113","DOIUrl":"https://doi.org/10.1109/JSTARS.2025.3602113","url":null,"abstract":"Leaf area index (LAI) is a key parameter for reflecting crop canopy structure and growth status, and crucial for precision agricultural management. Due to high acquisition costs and limitations of obtaining ground observation data, the radiative transfer models are widely used to generate simulated spectral data for inversion models. However, neglecting variations caused by factors like irrigation, fertilization, and soil differences may lead to some errors in inversion results. To address this problem, we proposed a stratified hybrid inversion method based on growth heterogeneity (GH-SHI). First, a simple linear iterative clustering algorithm was applied for superpixel-level image segmentation. Then, a growth status index was proposed to characterize the wheat growth heterogeneity and was used to classify the segmented regions into five winter wheat growth stages. For each category, simulated samples were generated using the coupled PROSPECT+SAIL (PROSAIL) radiative transfer model with the stratified parameters ranges to train BP neural network models and obtain LAI inversion results. Results show that the GH-SHI method achieves higher LAI inversion accuracy on the measured site data compared with the traditional hybrid method, with about a 50% reduction in RMSE (from 0.86 to 0.40), an improvement in <italic>R</i><inline-formula><tex-math>$^{2}$</tex-math></inline-formula> by 0.48 (from 0.13 to 0.61), and an increase in the correlation coefficient <italic>R</i> by 0.2 (from 0.64 to 0.84). Meanwhile, the correlation between the normalized difference vegetation index (NDVI) and LAI was used to further validate the effectiveness of the proposed method. The results show that, compared to the traditional approach, the LAI estimated by the GH-SHI method exhibits a stronger correlation with NDVI, further confirming the reliability of its inversion performance. In summary, the proposed method not only improves the accuracy of LAI inversion, but also demonstrates adaptability to different growth stages and diverse crop conditions, showing potential for application in large-scale, site-free regions.","PeriodicalId":13116,"journal":{"name":"IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing","volume":"18 ","pages":"21578-21592"},"PeriodicalIF":5.3,"publicationDate":"2025-08-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11134771","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145036934","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A Joint Real- and Complex-Valued Network for Classification of Pol(In)SAR Images","authors":"Yaobin Ma;Hossein Aghababaei;Ling Chang;Xiaohua Deng;Jingbo Wei","doi":"10.1109/JSTARS.2025.3602161","DOIUrl":"https://doi.org/10.1109/JSTARS.2025.3602161","url":null,"abstract":"Synthetic aperture radar (SAR) systems capture both amplitude and phase information, producing complex-valued images widely used in Earth observation applications. Among SAR modalities, polarimetric SAR and polarimetric interferometric SAR systems leverage polarimetric and interferometric information, typically represented by a data coherency matrix comprising real-valued diagonal elements and complex-valued off-diagonal elements. Existing deep learning approaches process the entire coherency matrix either through purely real-valued or purely complex-valued networks, which fails to fully exploit its heterogeneous structure. In this article, we propose a structurally decoupled modeling strategy for coherency matrices, which explicitly separates and processes diagonal and off-diagonal components based on their distinct structural roles. A real-valued network is employed to model the diagonal scattering power components, while a complex-valued network is used to capture the cross-channel correlation, orientation, and coherence structures from the off-diagonal components. This design aligns well with the inherent organization of the SAR coherency matrix, enabling more targeted and effective feature learning. The extracted real and complex features are subsequently fused via a cross-domain enhancement fusion block to achieve robust representation learning. Experiments on DLR’s FSAR dataset demonstrate that the proposed method consistently outperforms six state-of-the-art techniques across three different SAR modalities, achieving superior performance in both quantitative accuracy and qualitative robustness.","PeriodicalId":13116,"journal":{"name":"IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing","volume":"18 ","pages":"22256-22270"},"PeriodicalIF":5.3,"publicationDate":"2025-08-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11134815","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145051046","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}