ISPRS Journal of Photogrammetry and Remote Sensing最新文献

{"title":"Unveiling diurnal aerosol layer height variability from space using deep learning","authors":"Yulong Fan ,&nbsp;Lin Sun ,&nbsp;Zhihui Wang ,&nbsp;Shulin Pang ,&nbsp;Jing Wei","doi":"10.1016/j.isprsjprs.2025.08.021","DOIUrl":"10.1016/j.isprsjprs.2025.08.021","url":null,"abstract":"<div><div>The vertical distribution of aerosols is crucial for extensive climate and environment studies but is severely constrained by the limited availability of ground-based observations and the low spatiotemporal resolutions of Lidar satellite measurements. Multi-spectral passive satellites offer the potential to address these gaps by providing large-scale, high-temporal-resolution observations, making them a promising tool for enhancing current aerosol vertical distribution data. However, traditional methods, which rely heavily on physical assumptions and prior knowledge, often struggle to deliver robust and accurate aerosol vertical profiles. Thus, we develop a novel retrieval framework that combines two advanced deep-learning models, locally-feature-focused Transformer and globally-feature-focused Fully Connected Neural Network (FCNN), referred to as TF-FCNN, to estimate hourly aerosol distributions at different heights (i.e., 0.01–1 km, 1–2 km, and 2–3 km) with 2-km spatial resolution, using multi-source satellite data, including Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observations (CALIPSO), Himawari-8 and Moderate Resolution Imaging Spectroradiometer (MODIS). This hybrid framework is thoroughly analyzed using an eXplainable Artificial Intelligence (XAI)-based SHapley Additive exPlanations (SHAP) approach, which reveals that shortwave bands and brightness temperature are the most influential features, contributing approximately 63 % to the model predictions. Validation results demonstrate that the model provides reliable hourly aerosol vertical distributions across different heights in Australia, achieving high overall sample-based cross-validation coefficients of determination (CV-R²) ranging from 0.81 to 0.90 (average = 0.88). Our hourly retrievals indicate higher aerosol loadings at lower altitudes (0.01–1 km) than higher ones (1–2 km and 2–3 km) in most areas, likely due to significant anthropogenic and natural emissions from the ground. Furthermore, we observe substantial increases in aerosol concentrations over time and enhanced diurnal variations across altitudes during highly polluted cases, including urban haze and wildfires. These unique insights into the spatial distribution of aerosol vertical layers are crucial for effective air pollution control and management.</div></div>","PeriodicalId":50269,"journal":{"name":"ISPRS Journal of Photogrammetry and Remote Sensing","volume":"229 ","pages":"Pages 211-222"},"PeriodicalIF":12.2,"publicationDate":"2025-08-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144908140","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Cross-modal 2D-3D feature matching: simultaneous local feature description and detection across images and point clouds","authors":"Wei Ma ,&nbsp;Yucheng Huang ,&nbsp;Shengjun Tang ,&nbsp;Xianwei Zheng ,&nbsp;Zhen Dong ,&nbsp;Liang Ge ,&nbsp;Jianping Pan ,&nbsp;Qingquan Li ,&nbsp;Bing Wang","doi":"10.1016/j.isprsjprs.2025.08.016","DOIUrl":"10.1016/j.isprsjprs.2025.08.016","url":null,"abstract":"<div><div>Establishing correspondences between 2D images and 3D models is essential for precise 3D modeling and accurate positioning. However, widely adopted techniques for aligning 2D images with 3D features heavily depend on dense 3D reconstructions, which not only incur significant computational demands but also tend to exhibit reduced accuracy in texture-poor environments. In this study, we propose a novel method that combines local feature description and detection to enable direct and automatic alignment of 2D images with 3D models. Our approach utilizes a twin convolutional network architecture to process images and 3D data, generating respective feature maps. To address the non-uniform distribution of pixel and spatial point densities, we introduce an ultra-wide perception mechanism to expand the receptive field of image convolution kernels. Next, we apply a non-local maximum suppression criterion to concurrently evaluate the salience of pixels and 3D points. Additionally, we design an adaptive weight optimization loss function that dynamically guides learning objectives toward sample similarity. We rigorously validate our approach on multiple datasets, and our findings demonstrate successful co-extraction of cross-modal feature points. Through comprehensive 2D-3D feature matching experiments, we benchmark our method against several state-of-the-art techniques from recent literature. The results show that our method outperforms nearly all evaluated metrics, underscoring its effectiveness.</div></div>","PeriodicalId":50269,"journal":{"name":"ISPRS Journal of Photogrammetry and Remote Sensing","volume":"229 ","pages":"Pages 155-169"},"PeriodicalIF":12.2,"publicationDate":"2025-08-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144903652","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Scale-aware co-visible region detection for image matching","authors":"Xu Pan ,&nbsp;Zimin Xia ,&nbsp;Xianwei Zheng","doi":"10.1016/j.isprsjprs.2025.08.015","DOIUrl":"10.1016/j.isprsjprs.2025.08.015","url":null,"abstract":"<div><div>Matching images with significant scale differences remains a persistent challenge in photogrammetry and remote sensing. The scale discrepancy often degrades appearance consistency and introduces uncertainty in keypoint localization. While existing methods address scale variation through scale pyramids or scale-aware training, matching under significant scale differences remains an open challenge. To overcome this, we address the scale difference issue by detecting co-visible regions between image pairs and propose <strong>SCoDe</strong> (<strong>S</strong>cale-aware <strong>Co</strong>-visible region <strong>De</strong>tector), which both identifies co-visible regions and aligns their scales for highly robust, hierarchical point correspondence matching. Specifically, SCoDe employs a novel Scale Head Attention mechanism to map and correlate features across multiple scale subspaces, and uses a learnable query to aggregate scale-aware information of both images for co-visible region detection. In this way, correspondences can be established in a coarse-to-fine hierarchy, thereby mitigating semantic and localization uncertainties. Extensive experiments on three challenging datasets demonstrate that SCoDe outperforms state-of-the-art methods, improving the precision of a modern local feature matcher by 8.41%. Notably, SCoDe shows a clear advantage when handling images with drastic scale variations. Code is publicly available at <span><span>github.com/Geo-Tell/SCoDe</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":50269,"journal":{"name":"ISPRS Journal of Photogrammetry and Remote Sensing","volume":"229 ","pages":"Pages 122-137"},"PeriodicalIF":12.2,"publicationDate":"2025-08-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144895084","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"CGSL: Commonality graph structure learning for unsupervised multimodal change detection","authors":"Jianjian Xu ,&nbsp;Tongfei Liu ,&nbsp;Tao Lei ,&nbsp;Hongruixuan Chen ,&nbsp;Naoto Yokoya ,&nbsp;Zhiyong Lv ,&nbsp;Maoguo Gong","doi":"10.1016/j.isprsjprs.2025.08.010","DOIUrl":"10.1016/j.isprsjprs.2025.08.010","url":null,"abstract":"<div><div>Multimodal change detection (MCD) has attracted a great deal of attention due to its significant advantages in processing heterogeneous remote sensing images (RSIs) from different sensors (e.g., optical and synthetic aperture radar). The major challenge of MCD is that it is difficult to acquire the changed areas by directly comparing heterogeneous RSIs. Although many MCD methods have made important progress, they are still insufficient in capturing the modality-independence complex structural relationships in the feature space of heterogeneous RSIs. To this end, we propose a novel commonality graph structure learning (CGSL) for unsupervised MCD, which aims to extract potential commonality graph structural features between heterogeneous RSIs and directly compare them to detect changes. In this study, heterogeneous RSIs are first segmented and constructed as superpixel-based heterogeneous graph structural data consisting of nodes and edges. Then, the heterogeneous graphs are input into the proposed CGSL to capture the commonalities of graph structural features with modality-independence. The proposed CGSL consists of a Siamese graph encoder and two graph decoders. The Siamese graph encoder maps heterogeneous graphs into a shared space and effectively extracts potential commonality in graph structural features from heterogeneous graphs. The two graph decoders reconstruct the mapped node features as original node features to maintain consistency with the original graph features. Finally, the changes between heterogeneous RSIs can be detected by measuring the differences in commonality graph structural features using the mean squared error. In addition, we design a composite loss with regularization to guide CGSL in effectively excavating the potential commonality graph structural features between heterogeneous graphs in an unsupervised learning manner. Extensive experiments on seven MCD datasets show that the proposed CGSL outperforms the existing state-of-the-art methods, demonstrating its superior performance in MCD. The code will be available at <span><span>https://github.com/TongfeiLiu/CGSL-for-MCD</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":50269,"journal":{"name":"ISPRS Journal of Photogrammetry and Remote Sensing","volume":"229 ","pages":"Pages 92-106"},"PeriodicalIF":12.2,"publicationDate":"2025-08-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144895082","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Generalization of point-to-point matching for rigorous optimization in kinematic laser scanning","authors":"Aurélien Brun ,&nbsp;Jakub Kolecki ,&nbsp;Muyan Xiao ,&nbsp;Luca Insolia ,&nbsp;Elmar V. van der Zwan ,&nbsp;Stéphane Guerrier ,&nbsp;Jan Skaloud","doi":"10.1016/j.isprsjprs.2025.08.011","DOIUrl":"10.1016/j.isprsjprs.2025.08.011","url":null,"abstract":"<div><div>In the scope of rigorous sensor fusion in kinematic laser scanning, we present a qualitative improvement of an automated retrieval method of lidar-to-lidar 3D correspondences in terms of accuracy and speed, where correspondences are locally refined shifts derived from learning based descriptors matching. These improvements are shared through an open implementation. We evaluate their impact in three, fundamentally different laser scanning scenarios (sensors and platforms) without adaptation: airborne (helicopter), mobile (car) and handheld (without GNSS). The impact of precise correspondences improves the point cloud georeferencing/registration 2 to 10 times with respect to previously described and/or industrial standards, depending on the setup, without adaptation to a particular scenario. This represents a potential to enhance the accuracy and reliability of kinematic laser scanning in different environments, whether satellite positioning is available or not, and irrespectively of the nature of the lidars (i.e. including single-beam linear or oscillating sensors).</div></div>","PeriodicalId":50269,"journal":{"name":"ISPRS Journal of Photogrammetry and Remote Sensing","volume":"229 ","pages":"Pages 107-121"},"PeriodicalIF":12.2,"publicationDate":"2025-08-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144895083","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Polarization-Guided unsupervised convection networks for marine velocity field recovery","authors":"Yang Guo ,&nbsp;Naifu Yao ,&nbsp;Xi Lin ,&nbsp;Ning Li ,&nbsp;Yongqiang Zhao ,&nbsp;Seong G. Kong","doi":"10.1016/j.isprsjprs.2025.08.012","DOIUrl":"10.1016/j.isprsjprs.2025.08.012","url":null,"abstract":"<div><div>Accurate flow field measurement in the marine environment is crucial for promoting innovative development of ocean engineering. However, the limited concentration of deployable tracer particles and the complexities of marine environments often lead to unreliable flow field measurements. To address these challenges, we propose a marine environment flow field measurement system under a polarization optical framework. The proposed system utilizes the locally smooth characteristics of flow fields by designing an unsupervised convection network architecture to optimize the velocity field from sparse point clouds. Additionally, a tracer particle polarization feature discriminator is introduced to mitigate the interference from ghost particles. To support the system, a polarized light field sensor is developed to simultaneously capture three-dimensional and polarization information. The system is validated on both simulated and real-world datasets. Compared to existing studies confined to controlled laboratory conditions, the proposed system significantly enhances the applicability of particle tracking velocimetry technology in uncontrolled, complex marine environments. Quantitative evaluations demonstrate that our system achieves an EPE3D/m of 0.027, outperforming the state-of-the-art GotFlow3D method with 0.067. The paper resources can be viewed at <span><span>https://github.com/polwork</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":50269,"journal":{"name":"ISPRS Journal of Photogrammetry and Remote Sensing","volume":"229 ","pages":"Pages 17-31"},"PeriodicalIF":12.2,"publicationDate":"2025-08-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144863699","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"PAMSNet: A point annotation-driven multi-source network for remote sensing semantic segmentation","authors":"Yuanhao Zhao ,&nbsp;Mingming Jia ,&nbsp;Genyun Sun ,&nbsp;Aizhu Zhang","doi":"10.1016/j.isprsjprs.2025.07.035","DOIUrl":"10.1016/j.isprsjprs.2025.07.035","url":null,"abstract":"<div><div>Multi-source data semantic segmentation has proven to be an effective means of improving classification accuracy in remote sensing. With the rapid development of deep learning, the demand for large amounts of high-quality labeled samples has become a major bottleneck, limiting the broader application of these techniques. Weakly supervised learning has attracted increasing attention by reducing annotation costs. However, existing weakly supervised methods often suffer from limited accuracy. Effectively exploiting complementary information from multi-source remote sensing data using only a small number of labeled points remains a significant challenge. In this paper, we propose a novel architecture, named Point Annotation- Driven Multi-source Segmentation Network (PAMSNet), which leverages point annotations to effectively capture and integrate complementary features from multi-source remote sensing data. PAMSNet includes a Multi-source Feature Encoder and a Cross-Resolution Feature Integration (CRFI) module. The Multi-source Feature Encoder captures complementary global and local features using lightweight convolutional Global-Local Multi-source (GLMS) modules. And the boundary and spectral detail robustness are improved through Spectral-Edge Enhancement (SEE) modules, which effectively mitigate the impact of noise on segmentation accuracy. The CRFI module replaces conventional decoding structures by combining convolutional and Transformer mechanisms, enabling efficient cross-scale feature integration and improving the ability to identify multi-scale objects with reduced computational demands. Extensive experiments on the Vaihingen, WHU-IS, and WHU-OPT-SAR datasets validate the effectiveness of PAMSNet for weakly supervised multi-source segmentation as well as the validity of the proposed module. PAMSNet achieves state-of-the-art performance, with MIoU improvements of 2.4%, 2.1%, and 3.16% on three datasets, using only 0.01% point annotations. Additionally, PAMSNet can effectively balance the performance as well as the operational efficiency of the model compared to existing methods, which further promotes the application of deep learning in remote sensing image mapping.</div></div>","PeriodicalId":50269,"journal":{"name":"ISPRS Journal of Photogrammetry and Remote Sensing","volume":"229 ","pages":"Pages 1-16"},"PeriodicalIF":12.2,"publicationDate":"2025-08-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144863822","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"FineCrop: Mapping fine-grained crops using class-aware feature decoupling and parcel-aware class rebalancing with Sentinel-2 time series","authors":"Lei Lei ,&nbsp;Xinyu Wang ,&nbsp;Yanfei Zhong ,&nbsp;Liangpei Zhang","doi":"10.1016/j.isprsjprs.2025.07.041","DOIUrl":"10.1016/j.isprsjprs.2025.07.041","url":null,"abstract":"<div><div>Fine-grained crop mapping refers to the precise differentiation of all crop types within an area, encompassing major classes (e.g., staple crops, to cash crops, to garden fruits, etc.) and their subclasses (e.g., wheat, barley, maize of staple crops). Fine-grained crop mapping is crucial for precise agriculture management. However, compared with staple crop mapping, fine-grained crop mapping faces more challenges: (1) the extremely similar phenological characteristics between crop subcategories, which could lead to the difficulty in extracting discriminative representation; (2) the imbalanced class distribution, which could lead to the bias of the model toward the head class, finally causing severe misclassification. In this paper, we proposed a novel framework for fine-grained crop mapping, termed FineCrop, by using class-aware feature decoupling (CFD) branch and parcel-aware class rebalancing (PCR) branch. Specifically, CFD was inspired by the “divide and conquer” theory and designed to learn the detailed and independent features of each crop type and to solve the phenological similarity. PCR was inspired by the data aggregation and designed to use a class-aware factor at parcel unit to solve the bias of classifier caused by the imbalanced data distribution. To evaluate FineCrop, we have built a fine-grained crops mapping dataset, termed FineCropSet by matching Sentinel-2 Level-2A product that has undergone radiometric and geometric correction with labels extracted from the EuroCrops. FineCropSet contains 138 crop types covering the North Rhine-Westphalia, the south of Slovakia, and the Netherlands of different years. The results showed that FineCrop can improve the overall accuracy of popular deep learning models for temporal satellite imagery by 5.83 %, 1.42 %, 0.89 % for three study areas respectively (p-value less than 0.05) verified by paired <em>t</em>-test, confirming the substantial improvement of FineCrop for fine-grained crop mapping. The ablation experiment results revealed that FineCrop could reduce the class imbalance by 5 and 18 times and extract the detailed features in parcel from edge to center. We believe the proposed method is promising for large-scale crop mapping particularly when dealing with crops with similar phenological characteristics and imbalanced distribution, leading to more accurate crop resources inventory. The source code and data are available: <span><span>https://github.com/LL0912/FineCrop</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":50269,"journal":{"name":"ISPRS Journal of Photogrammetry and Remote Sensing","volume":"228 ","pages":"Pages 785-803"},"PeriodicalIF":12.2,"publicationDate":"2025-08-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144830657","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Multi-level Priors-Guided Diffusion-based Remote Sensing Image Super-Resolution","authors":"Lijing Lu ,&nbsp;Zhou Huang ,&nbsp;Yi Bao ,&nbsp;Lin Wan ,&nbsp;Zhihang Li","doi":"10.1016/j.isprsjprs.2025.07.020","DOIUrl":"10.1016/j.isprsjprs.2025.07.020","url":null,"abstract":"<div><div>Recently, diffusion models have achieved advancements in natural image super-resolution (SR) tasks, overcoming some issues posed by traditional approaches, e.g., performance limitations in CNN-based and Transformer-based approaches, as well as instable training and mode collapse in GAN. However, despite these advancements, existing diffusion-based SR methods fail to perform well for remote sensing images. Current diffusion-based super-resolution techniques face two key challenges: (1) A jeopardy to the generative prior arises due to the necessity of training from scratch, which can lead to suboptimal performance. (2) A loss of fidelity occurs due to the limited priors in SR models, which only take the low-resolution image as input. To deal with these challenges, we introduce a Multi-level Priors-Guided Diffusion-based Remote Sensing Image Super-Resolution Model (DLMSR) approach. In particular, we utilize a pre-trained stable diffusion model to maintain the generative prior captured in synthesis models, resulting in more stable and detailed outcomes. Furthermore, to establish comprehensive priors, we incorporate multimodal large language models (MLLMs) to capture diverse priors such as texture and content priors. Additionally, we introduce category priors by employing a category classifier to offer global and concise signals for precise reconstruction. Then, we devise a cascade prior fusion module and a class-aware encoder to integrate rich priors into the diffusion model. DLMSR is extensively evaluated on four publicly available remote sensing datasets, including AID, DOTA, DIOR, and NWPU-RESISC45, demonstrating consistent advantages over representative state-of-the-art methods. In particular, compared with StableSR, DLMSR achieves an average increase of 0.29 dB in PSNR and a decrease of 1.93 in FID across three simulated benchmarks, indicating enhanced reconstruction fidelity and perceptual quality. The source code and dataset links are publicly available at: <span><span>https://github.com/lijing28/DLMSR.git</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":50269,"journal":{"name":"ISPRS Journal of Photogrammetry and Remote Sensing","volume":"228 ","pages":"Pages 756-770"},"PeriodicalIF":12.2,"publicationDate":"2025-08-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144810215","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Unmixing frequency features for DEM super resolution","authors":"Zhuwei Wen, He Chen, Xianwei Zheng","doi":"10.1016/j.isprsjprs.2025.07.039","DOIUrl":"https://doi.org/10.1016/j.isprsjprs.2025.07.039","url":null,"abstract":"DEM super-resolution (SR) has recently been advanced by deep learning. The focus of existing works is mainly on the employment of various terrain constraints to force the general deep SR models to adapt to DEM data. However, we found that they leave a fundamental issue of terrain pattern confusion caused by the mixed frequency feature learning of deep neural networks, which leads to an inherent trade-off between the reconstruction of fundamental structures and the preservation of fine-grained terrain details. In this study, we propose a novel dual-frequency feature learning network (DuffNet) for high quality DEM super-resolution. The core idea of DuffNet is to directly learn the mapping relationship between low-resolution (LR) and high-resolution (HR) DEMs with meaningful frequency features, rather than the mixed convolutional features extracted from raw DEMs. Specifically, DuffNet deploys a dual-branch structure with a dedicatedly designed dual-frequency loss to enable the learning of high- and low-frequency features under the supervision of input HR DEM. An adaptive elevation amplitude refiner (AEAR) is then developed to dynamically adjust and optimize the amplitudes of the initial HR DEM synthesized by the integration of learned low-frequency and high-frequency terrain components. Extensive experiments conducted on TFASR30, Pyrenees, Tyrol, and the challenging TFASR30to10 datasets show that DuffNet can achieve state-of-the-art performance, outperforming other SoTA methods such as TTSR and CDEM by 19% and 29% respectively in RMSE-Elevation on the TFASR30to10 dataset. The dataset and source code are available at: <ce:inter-ref xlink:href=\"https://github.com/Geo-Tell/DuffNet\" xlink:type=\"simple\">https://github.com/Geo-Tell/DuffNet</ce:inter-ref>.","PeriodicalId":50269,"journal":{"name":"ISPRS Journal of Photogrammetry and Remote Sensing","volume":"42 1","pages":""},"PeriodicalIF":12.7,"publicationDate":"2025-08-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144898133","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}