Remote Sensing

20 pages, 749 KiB

Open AccessArticle

Direction of Arrival (DOA) Estimation Using a Deep Unfolded Learned Iterative Shrinkage Thresholding Algorithm (LISTA) Network in a Non-Uniform Metasurface

by Xinyi Niu, Xiaolong Su, Lida He and Guanchao Chen

Remote Sens. 2025, 17(7), 1253; https://doi.org/10.3390/rs17071253 (registering DOI) - 1 Apr 2025

Abstract

This paper proposes a novel method for Direction of Arrival (DOA) estimation using a deep unfolded LISTA network in a non-uniform metasurface. Traditional DOA estimation methods often face challenges such as limited accuracy, high computational complexity, and poor adaptability to complex signal environments. [...] Read more.

This paper proposes a novel method for Direction of Arrival (DOA) estimation using a deep unfolded LISTA network in a non-uniform metasurface. Traditional DOA estimation methods often face challenges such as limited accuracy, high computational complexity, and poor adaptability to complex signal environments. To address these issues, we optimize a non-uniform metasurface array to reduce hardware costs and mutual coupling effects while enhancing resolution. Additionally, a deep unfolded Learned Iterative Shrinkage Thresholding Algorithm (LISTA) network is constructed by transforming Iterative Shrinkage Thresholding Algorithm (ISTA) iterative steps into trainable neural network layers, combining model-driven logic with data-driven parameter optimization. Simulation results prove that this method enhances higher precision and reduces computational complexity in comparison with traditional algorithms, especially under low SNR conditions. Furthermore, the method exhibits greater generalization ability, making it a reliable approach for high-precision DOA estimation in practical applications. Full article

(This article belongs to the Special Issue Array and Signal Processing for Radar)

17 pages, 1909 KiB

Open AccessArticle

Solar-Induced Fluorescence as Indicator of Downy Oak and the Influence of Some Environmental Variables at the End of the Growing Season

by Antoine Baulard, Jean-Philippe Mevy, Irène Xueref-Remy, Ilja Marco Reiter, Tommaso Julitta and Franco Miglietta

Remote Sens. 2025, 17(7), 1252; https://doi.org/10.3390/rs17071252 (registering DOI) - 1 Apr 2025

Abstract

In the context of global warming, which is mainly due to the increasing atmospheric concentration of carbon dioxide, the prediction of climate change requires a good assessment of the involvement of vegetation in the global carbon cycle. In particular, determining when vegetative activity [...] Read more.

In the context of global warming, which is mainly due to the increasing atmospheric concentration of carbon dioxide, the prediction of climate change requires a good assessment of the involvement of vegetation in the global carbon cycle. In particular, determining when vegetative activity ceases in deciduous forests remains a great challenge. Remote sensing of solar-induced fluorescence (SIF) has been considered as a potential proxy for ecosystem photosynthesis and, therefore, a relevant indicator of the end of the vegetation period as compared to other vegetation indices, such as EVI (Enhanced Vegetation Index) and NDVI (Normalized Difference Vegetation Index). However, many challenges remain to be addressed due to the lack of knowledge of the response of SIF at different time scales, different species and different environmental conditions. The aim of this study was to explore the diurnal and seasonal variations in the SIFA and SIFB signals in a pubescent oak forest undergoing senescence. We show that apparent SIFA yield may be considered an earlier indicator of the end of the vegetation period compared to NDVI, which primarily reflects the ratio of SIFB/SIFA. The apparent SIFA yield signal was positively and highly correlated with PRI (Photochemical Reflectance Index), EVI and NDVI. Air contents in CO₂ and O₃ were similarly significantly correlated to SIFs emission but only during the growth phase of the phenology of Q. pubescens. At the seasonal scale, the results show that SIF variations were mainly driven by variations in PAR, air VPD and temperature. A higher dependence of the SIF signal on these last three variables was observed at the diurnal scale through Pearson correlation coefficients, which were greater than seasonal ones. Full article

(This article belongs to the Section Ecological Remote Sensing)

28 pages, 13136 KiB

Open AccessArticle

Fine-Tuning-Based Transfer Learning for Building Extraction from Off-Nadir Remote Sensing Images

by Bipul Neupane, Jagannath Aryal and Abbas Rajabifard

Remote Sens. 2025, 17(7), 1251; https://doi.org/10.3390/rs17071251 (registering DOI) - 1 Apr 2025

Abstract

Building extraction—needed for urban planning and monitoring—is affected by the misalignment between labels and off-nadir remote sensing imagery. A computer vision approach to teacher–student learning between large–noisy and small–clean data has been introduced as a solution, but with limited accuracy and efficiency. This [...] Read more.

Building extraction—needed for urban planning and monitoring—is affected by the misalignment between labels and off-nadir remote sensing imagery. A computer vision approach to teacher–student learning between large–noisy and small–clean data has been introduced as a solution, but with limited accuracy and efficiency. This paper proposes fine-tuning-based transfer learning (FTL) to adapt a pre-trained model from a noisy source to a clean target dataset, improving segmentation accuracy in off-nadir images. A standardized experimental framework is developed with three new building datasets containing large–noisy and small–clean image–label pairs of multiple spatial resolutions. These datasets cover a range of building types, from low-rise to skyscrapers. Additionally, this paper presents one of the most extensive benchmarking efforts in teacher–student learning for building extraction from off-nadir images. Results demonstrate that FTL outperforms the existing methods with higher F1 scores—0.943 (low-rise), 0.868 (mid-rise), 0.912 (high-rise), and 0.697 (skyscrapers)—and higher computational efficiency. A notable gain in mean difference is observed in taller buildings from complex urban environments. The proposed method, datasets, and benchmarking framework provide a robust foundation for accurate building extraction and broader remote sensing applications. Full article

(This article belongs to the Special Issue Applications of AI and Remote Sensing in Urban Systems II)

15 pages, 1218 KiB

Open AccessCommunication

Microphysical Characteristics of Convective and Stratiform Precipitation Generated at Different Life Stages of Precipitating Cloud in the Pre-Summer Rainy Season in South China

by Jiayan Yang, Yunying Li, Xiong Hu, Zhiwei Zhang and Xiongwei Kou

Remote Sens. 2025, 17(7), 1250; https://doi.org/10.3390/rs17071250 (registering DOI) - 1 Apr 2025

Abstract

This study uses GPM DPR and Himawari-8 cloud-top infrared data to classify the precipitating cloud (PC) into three life stages: developing, mature, and dissipating. Based on GPM DPR data from April to June 2018–2022, this research investigates the microphysical features of convective and [...] Read more.

This study uses GPM DPR and Himawari-8 cloud-top infrared data to classify the precipitating cloud (PC) into three life stages: developing, mature, and dissipating. Based on GPM DPR data from April to June 2018–2022, this research investigates the microphysical features of convective and stratiform precipitation over South China. The precipitation generated by the developing stage of the PC contains the largest proportion of convective precipitation, the largest precipitation area in the mature stage of PC, and the smallest precipitation area with the lowest convective precipitation proportion in the dissipating stage of the PC. For stratiform precipitation generated by the developing PC, the height of 0 °C level is marginally above the top height of Bright Band (BB), with both heights aligning in altitude during the mature and dissipating stages of the PC. The mass-weighted mean diameter (D_m) peaks at 1.2 mm below the BB, and near-surface D_m is positively correlated with the storm top height. For convective precipitation, raindrops with D_m of 1.9 mm and those exceeding 3.0 mm predominate. Notably, the near-surface D_m shows a positive correlation with storm top height, with the correlation coefficient for convective precipitation being greater than that for stratiform precipitation. Significantly, the average liquid and non-liquid water paths are larger in the dissipating stage as compared to the developing stage for both precipitation types. These findings suggest enhanced precipitation efficiency in South China and underscore the critical importance of stage-specific analyses in comprehending precipitating cloud microphysics. Full article

(This article belongs to the Special Issue Synergetic Remote Sensing of Clouds and Precipitation II)

19 pages, 6858 KiB

Open AccessArticle

Application Possibilities of Orthophoto Data Based on Spectral Fractal Structure Containing Boundary Conditions

by József Berke

Remote Sens. 2025, 17(7), 1249; https://doi.org/10.3390/rs17071249 (registering DOI) - 1 Apr 2025

Abstract

The self-similar structure-based analysis of digital images offers many new practical possibilities. The fractal dimension is one of the most frequently measured parameters if we want to use image data in measurable analyses in metric spaces. In practice, the fractal dimension can be [...] Read more.

The self-similar structure-based analysis of digital images offers many new practical possibilities. The fractal dimension is one of the most frequently measured parameters if we want to use image data in measurable analyses in metric spaces. In practice, the fractal dimension can be measured well in simple files containing only image data. In the case of complex image data structures defined in different metric spaces, their measurement in metric space encounters many difficulties. In this work, we provide a practical solution for the measurement of ortho-photos—as complex image data structures—based on the spectral fractal structure based on boundary conditions (height, time, and temperature), presenting the further development of the related theoretical foundations. We will discuss the optimal flight altitude determination in detail through practical examples. For this, in addition to the structural measurements on the images, we also use the well-known image entropy in information theory. The data obtained in this way can facilitate the optimal UAS operation execution that best suits further image processing tasks (e.g., classification, segmentation, and index analysis). Full article

(This article belongs to the Special Issue Image Processing from Aerial and Satellite Imagery)

► Show Figures

Figure 1

17 pages, 6898 KiB

Open AccessArticle

Change Patterns of Ecological Vulnerability and Its Dominant Factors in Mongolia During 2000–2022

by Jing Han, Bing Guo, Lizhi Pan, Baomin Han and Tianhe Xu

Remote Sens. 2025, 17(7), 1248; https://doi.org/10.3390/rs17071248 (registering DOI) - 1 Apr 2025

Abstract

Under global climate change, the ecological vulnerability issue in Mongolia has become increasingly severe. However, the change process of the ecological environment and the dominant driving factors in different periods and sub-regions of Mongolia are not clear. In this paper, we propose a [...] Read more.

Under global climate change, the ecological vulnerability issue in Mongolia has become increasingly severe. However, the change process of the ecological environment and the dominant driving factors in different periods and sub-regions of Mongolia are not clear. In this paper, we propose a new ecological vulnerability index for Mongolia using MODIS data, combined with the Geographical Detector and the gravity center model, to reveal the spatiotemporal changes and driving mechanisms of ecological vulnerability in Mongolia from 2000 to 2022. The results show the following: (1) the newly proposed remote sensing ecological vulnerability index has high applicability in ecosystems mainly in Mongolia, with an accuracy rate of 89.39%; (2) Mongolia belongs to the category of moderate vulnerability, with an average ecological vulnerability index of 1.57, and the center of vulnerability is shifting toward the southwest direction; (3) Tmax is the leading driving factor of ecological vulnerability in Mongolia, especially at high altitudes and in arid regions, where it directly affects vegetation growth, desertification, and water availability. The dominant interactive factors have shifted from Tmax ∩ Tmin to Tmin ∩ PRE, with PRE being the leading factor in the eastern, central, and southern regions of Mongolia, Tmax being the leading factor in the western region, and Tmin being the leading factor in the northwestern region. This study provides an index system for constructing the ecological vulnerability system in Mongolia and offers scientific references for the regional protection of the ecological environment in Mongolia. Full article

(This article belongs to the Special Issue Recent Advances in Remote Sensing of Soil Science)

► Show Figures

Figure 1

36 pages, 17840 KiB

Open AccessArticle

Advanced Building Detection with Faster R-CNN Using Elliptical Bounding Boxes for Displacement Handling

by Sejung Jung, Ahram Song, Kirim Lee and Won Hee Lee

Remote Sens. 2025, 17(7), 1247; https://doi.org/10.3390/rs17071247 (registering DOI) - 1 Apr 2025

Abstract

This study presents an enhanced Faster R-CNN framework that incorporates elliptical bounding boxes to significantly improve building detection in off-nadir imagery, effectively reducing severe geometric distortions caused by oblique sensor angles. Off-nadir imagery enhances architectural detail capture and reduces occlusions, but conventional bounding [...] Read more.

This study presents an enhanced Faster R-CNN framework that incorporates elliptical bounding boxes to significantly improve building detection in off-nadir imagery, effectively reducing severe geometric distortions caused by oblique sensor angles. Off-nadir imagery enhances architectural detail capture and reduces occlusions, but conventional bounding boxes, such as axis-aligned and rotated bounding boxes, often fail to localize buildings distorted by extreme perspectives. We propose a hybrid method integrating elliptical bounding boxes for curved structures and rotated bounding boxes for tilted buildings, achieving more precise shape approximation. In addition, our model incorporates a squeeze-and-excitation mechanism to refine feature representation, suppress background noise, and enhance object boundary alignment, leading to superior detection accuracy. Experimental results on the BONAI dataset demonstrate that our approach achieves a detection rate of 91.96%, significantly outperforming axis-aligned bounding boxes (65.75%) and rotated bounding boxes (87.13%) in detecting irregular and distorted buildings. By providing a highly robust and adaptable detection strategy, our approach establishes a new standard for accurate and shape-aware building recognition in off-nadir imagery, significantly improving the detection of distorted, rotated, and irregular structures. Full article

► Show Figures

Figure 1

35 pages, 12903 KiB

Open AccessReview

Perfect Labelling: A Review and Outlook of Label Optimization Techniques in Dynamic Earth Observation

by Sarah Hauser, Lena Augner and Andreas Schmitt

Remote Sens. 2025, 17(7), 1246; https://doi.org/10.3390/rs17071246 (registering DOI) - 1 Apr 2025

Abstract

Advances in Artificial Intelligence (AI) and Machine Learning (ML) have significantly enhanced the practice of Earth Observation (EO), enabling complex analyses such as land cover change detection, vegetation monitoring, and disaster response. However, while model architectures have matured, the refinement of reference data [...] Read more.

Advances in Artificial Intelligence (AI) and Machine Learning (ML) have significantly enhanced the practice of Earth Observation (EO), enabling complex analyses such as land cover change detection, vegetation monitoring, and disaster response. However, while model architectures have matured, the refinement of reference data remains a major challenge. Accurate and dynamic multi-temporal labelling is essential for capturing evolving ground conditions in high-dimensional EO datasets, yet key challenges persist, including spatiotemporal inconsistencies, heterogeneous data integration, and multi-resolution harmonization. Without robust preprocessing, reference labels may introduce biases, resulting in reduced model reliability and generalizability. This review tackles four core aspects of reference data preprocessing in EO: (i) essential steps for producing consistent and high-quality datasets, particularly for dynamic spatiotemporal data; (ii) best practices and guidelines that enable scalable and accurate workflows across diverse EO applications; (iii) introduction of the HELIX framework, a unified approach for standardizing, enhancing, and automating spatiotemporal label preprocessing; and (iv) a forward-looking discussion on the future of reference labels and features, including next-generation techniques for dynamic EO data integration. By synthesizing existing methodologies, highlighting emerging approaches, and addressing current gaps, this review underscores how well-engineered reference data are fundamental to advancing AI/ML-driven EO applications. Full article

(This article belongs to the Special Issue State-of-the-Art in Land Cover Classification and Mapping)

► Show Figures

Figure 1

21 pages, 8338 KiB

Open AccessArticle

The Predictive Skill of a Remote Sensing-Based Machine Learning Model for Ice Wedge and Visible Ground Ice Identification in Western Arctic Canada

by Qianyu Chang, Simon Zwieback and Aaron A. Berg

Remote Sens. 2025, 17(7), 1245; https://doi.org/10.3390/rs17071245 (registering DOI) - 1 Apr 2025

Abstract

Fine-scale maps of ground ice and related surface features are critical for permafrost-related modelling and management. However, such maps are lacking across almost the entire Arctic. Machine learning provides the potential to automate regional fine-scale ground ice mapping using remote sensing and topographic [...] Read more.

Fine-scale maps of ground ice and related surface features are critical for permafrost-related modelling and management. However, such maps are lacking across almost the entire Arctic. Machine learning provides the potential to automate regional fine-scale ground ice mapping using remote sensing and topographic data. Here, we evaluate the predictive skill of XGBoost models for identifying (1) ice wedge and (2) top-5m visible ground ice in the Tuktoyaktuk Coastlands. We find high predictive skill for ice wedge occurrence (ROC AUC = 0.95, macro F1 = 0.80), with the most important predictors being slope, distance to the coast, and probability of depression. The model accurately predicted regional and local trends in ice wedge occurrence, with an increase in ice wedge polygon (IWP) probability towards the coast and in poorly drained depressions. The model also captured IWP in well-drained uplands of the study area, including locations with poorly visible troughs not contained in the training data. Spatial transferability analyses highlight the regional variability of ice wedge probability, reflecting contrasting climatic and surface conditions. Conversely, the low predictive skill for visible ground ice (ROC AUC = 0.67, macro F1 = 0.53) is attributed to limitations in training data and weak associations with the remotely sensed predictors. The varying predictive accuracy highlights the importance of high-quality reference data and site-specific conditions for improving ground ice studies with data-driven modelling from remote sensing observations. Full article

(This article belongs to the Special Issue The Applications of Remote Sensing, Machine Learning and Deep Learning in Frozen Ground Regions)

► Show Figures

Figure 1

23 pages, 36274 KiB

Open AccessArticle

An Improved Machine Learning-Based Method for Unsupervised Characterisation for Coral Reef Monitoring in Earth Observation Time-Series Data

by Zayad AlZayer, Philippa Mason, Robert Platt and Cédric M. John

Remote Sens. 2025, 17(7), 1244; https://doi.org/10.3390/rs17071244 (registering DOI) - 1 Apr 2025

Abstract

This study presents an innovative approach to automated coral reef monitoring using satellite imagery, addressing challenges in image quality assessment and correction. The method employs Principal Component Analysis (PCA) coupled with clustering for efficient image selection and quality evaluation, followed by a machine [...] Read more.

This study presents an innovative approach to automated coral reef monitoring using satellite imagery, addressing challenges in image quality assessment and correction. The method employs Principal Component Analysis (PCA) coupled with clustering for efficient image selection and quality evaluation, followed by a machine learning-based cloud removal technique using an XGBoost model trained to detect land and cloudy pixels over water. The workflow incorporates depth correction using Lyzenga’s algorithm and superpixel analysis, culminating in an unsupervised classification of reef areas using KMeans. Results demonstrate the effectiveness of this approach in producing consistent, interpretable classifications of reef ecosystems across different imaging conditions. This study highlights the potential for scalable, autonomous monitoring of coral reefs, offering valuable insights for conservation efforts and climate change impact assessment in shallow marine environments. Full article

► Show Figures

Figure 1

23 pages, 5667 KiB

Open AccessArticle

Validating HF Radar Current Accuracy via Lagrangian Measurements and Radar-to-Radar Comparisons in Highly Variable Surface Currents

by Bartolomeo Doronzo, Michele Bendoni, Stefano Taddei, Angelo Boccacci and Carlo Brandini

Remote Sens. 2025, 17(7), 1243; https://doi.org/10.3390/rs17071243 (registering DOI) - 31 Mar 2025

Abstract

The validation of HF radar systems remains an area with significant scope for advancement, particularly in terms of linking data quality with system operational parameters, fully utilizing the potential of redundant data (e.g., overlapping radial measurements), and accurately capturing the spatiotemporal variability observed [...] Read more.

The validation of HF radar systems remains an area with significant scope for advancement, particularly in terms of linking data quality with system operational parameters, fully utilizing the potential of redundant data (e.g., overlapping radial measurements), and accurately capturing the spatiotemporal variability observed by independent devices, such as drifters. In this study, we conducted a large-scale Lagrangian measurement campaign in the Tuscan Archipelago, aimed at validating surface current data from the HF radar network. This radar network, a recent addition to the area, monitors an oceanographic region critical to Mediterranean dynamics. The validation was executed using different approaches: a Eulerian method, comparing the radial velocities measured by radar with drifter-derived velocities along radial directions; a Lagrangian method, contrasting the observed drifter trajectories with the synthetic virtual trajectories generated from radar-based flow fields; and radar-to-radar comparisons with the concurrent utilization of two radars in same point. Through fine-tuning of the quality control parameters and an analysis of the impact of different thresholds of such parameters, we assessed the radar’s ability to capture dynamic processes, identifying both strengths and limitations. Our results not only confirm the utility of HF radar in coastal monitoring but also provide a basis for improving calibration strategies, ultimately supporting more accurate, high-resolution radar observations in complex marine environments. Full article

(This article belongs to the Special Issue Remote Sensing of Ocean Surface Currents: Measurement, Validation and Applications)

► Show Figures

Figure 1

23 pages, 25882 KiB

Open AccessArticle

Robust Low-Sidelobe MIMO Dual-Function Radar–Communication Waveform Design

by Xuchen Liu, Yongjun Liu, Guisheng Liao, Hao Tang, Heming Wang and Xiaoyang Dong

Remote Sens. 2025, 17(7), 1242; https://doi.org/10.3390/rs17071242 (registering DOI) - 31 Mar 2025

Abstract

In multi-input–multi-output (MIMO) dual-function radar–communication (DFRC) systems, the inevitable amplitude–phase errors increase the sidelobe of transmit beampattern and distort the synthesized waveforms, which degrades both radar and communication performance. Due to this, a robust low-sidelobe MIMO DFRC waveform design method is proposed. Firstly, [...] Read more.

In multi-input–multi-output (MIMO) dual-function radar–communication (DFRC) systems, the inevitable amplitude–phase errors increase the sidelobe of transmit beampattern and distort the synthesized waveforms, which degrades both radar and communication performance. Due to this, a robust low-sidelobe MIMO DFRC waveform design method is proposed. Firstly, a DFRC transmit signal model based on the uncertainty sets of amplitude–phase errors is established. The robust low-sidelobe MIMO DFRC waveform design problem is then formulated. In this problem, the sidelobe of transmit beampattern is minimized with the constraints on the mutual interference and the desired waveforms. To decrease the computational complexity, an alternating direction method of multipliers (ADMM)-based waveform design method is proposed, and the convergence is proved. Finally, some simulation results are presented to validate the effectiveness of the proposed method. Full article

► Show Figures

Figure 1

13 pages, 3745 KiB

Open AccessCommunication

Seismo-Traveling Ionospheric Disturbances from the 2024 Hualien Earthquake: Altitude-Dependent Propagation Insights

by Zhiqiang Mao, Chieh-Hung Chen, Aisa Yisimayili, Jing Liu, Xuemin Zhang, Yang-Yi Sun, Yongxin Gao, Shengjia Zhang, Chuanqi Teng and Jianjun Zhao

Remote Sens. 2025, 17(7), 1241; https://doi.org/10.3390/rs17071241 (registering DOI) - 31 Mar 2025

Abstract

The propagation of seismo-traveling ionospheric disturbances (STIDs) is generally observed at one specific altitude layer. On 2 April 2024, a Mw 7.4 earthquake struck Hualien, which was the biggest earthquake since the 1999 Chi-Chi earthquake in the Taiwan region. In this study, a [...] Read more.

The propagation of seismo-traveling ionospheric disturbances (STIDs) is generally observed at one specific altitude layer. On 2 April 2024, a Mw 7.4 earthquake struck Hualien, which was the biggest earthquake since the 1999 Chi-Chi earthquake in the Taiwan region. In this study, a co-located vertical monitoring system combined with the observation of two horizontal layers in the ionosphere was utilized to study the STIDs associated with the Hualien earthquake. The vertical monitoring system can capture disturbances from the ground surface up to a height of ~350 km. In addition, changes in electric currents and the TEC (total electron content) at two horizontal layers, ~100 km and ~350 km, were monitored by permanent geomagnetic stations and a ground-based GNSS (global navigation satellite system) receivers network, respectively. The observations from this four-dimensional (4D) monitoring network show that the STIDs at a height of ~100 km associated with Rayleigh waves can propagate as far as 2000 km from the epicenter, while at an altitude of ~350 km, they can only propagate to about 1000 km. At an altitude of about 200 km, STIDs were also captured by a high-frequency Doppler sounder in a vertical monitoring system, which was consistent with the results in the geomagnetic field. The results from the 4D monitoring network suggest that the STIDs associated with Rayleigh waves exhibit different propagation ranges at various altitudes and prefer to propagate at low ionosphere layers. The vertical propagating waves typically only reach the bottom of the ionosphere and struggle to propagate to higher regions over long distances. Full article

(This article belongs to the Special Issue State of the Art of Geomagnetic/Electromagnetic Satellites: Science and Applications (Second Edition))

► Show Figures

Figure 1

23 pages, 2392 KiB

Open AccessArticle

MDFusion: Multi-Dimension Semantic–Spatial Feature Fusion for LiDAR–Camera 3D Object Detection

by Renzhong Qiao, Hao Yuan, Zhenbo Guan and Wenbo Zhang

Remote Sens. 2025, 17(7), 1240; https://doi.org/10.3390/rs17071240 (registering DOI) - 31 Mar 2025

Abstract

Accurate 3D object detection is becoming increasingly vital for the development of robust perception systems, particularly in applications such as autonomous driving vehicles and robotic systems. Many existing approaches rely on bird’s eye view (BEV) feature maps to facilitate multi-modal interaction, as BEV [...] Read more.

Accurate 3D object detection is becoming increasingly vital for the development of robust perception systems, particularly in applications such as autonomous driving vehicles and robotic systems. Many existing approaches rely on bird’s eye view (BEV) feature maps to facilitate multi-modal interaction, as BEV representations enable efficient operations. However, the inherent sparsity of LiDAR BEV features often leads to misalignment with the dense semantic information in camera images, resulting in suboptimal fusion quality and degraded detection performance, especially in complex and dynamic environments. To mitigate these issues, this paper proposes a novel multi-dimension semantic–spatial feature fusion (MDFusion) method that combines LiDAR and image features in 2D and 3D spaces. Specifically, image semantic features are extracted using the DeepLabV3 segmentation network, which captures rich contextual information and is aligned with LiDAR point cloud voxel features through a summation operation to achieve precise semantic fusion. Additionally, LiDAR BEV features are fused with downsampled image features in 2D space via concatenation and spatially adaptive dilated convolution. The mechanism dynamically adjusts to the spatial characteristics of the data, ensuring robust feature integration. Extensive experiments on the KITTI and ONCE datasets demonstrate that our method achieves competitive performance in complex scenes, significantly improving the multi-modal fusion quality and detection accuracy while maintaining computational efficiency. Full article

► Show Figures

Figure 1

19 pages, 6292 KiB

Open AccessArticle

EFCNet: Expert Feature-Based Convolutional Neural Network for SAR Ship Detection

by Zheng Chen, Yuxiang Zhang, Jing Bai and Biao Hou

Remote Sens. 2025, 17(7), 1239; https://doi.org/10.3390/rs17071239 (registering DOI) - 31 Mar 2025

Abstract

Due to the special properties of synthetic aperture radar (SAR) images, they are widely used in maritime applications, such as detecting ships at sea. To perform ship detection in SAR images, existing algorithms commonly utilize convolutional neural network (CNN). However, the challenges in [...] Read more.

Due to the special properties of synthetic aperture radar (SAR) images, they are widely used in maritime applications, such as detecting ships at sea. To perform ship detection in SAR images, existing algorithms commonly utilize convolutional neural network (CNN). However, the challenges in acquiring SAR images and the imaging noise hinder CNN in performing SAR ship-detection tasks. In this paper, we revisit the relationship between SAR expert features and network abstract features, and propose an expert-feature-based convolutional neural network (EFCNet). Specifically, we exploit the inherent physical properties of SAR images by manually extracting a range of expert features, including electromagnetic scattering, geometric structure, and grayscale statistics. These expert features are then adaptively integrated with abstract CNN features through a newly designed multi-source features association module, which improves the common CNN’s capability to recognize ship targets. Experiment results on the SSDD demonstrate that EFCNet outperforms general CNN approaches. Furthermore, EFCNet achieves comparable detection performance to baseline methods while utilizing only 70% of the data capacity, highlighting its efficiency. This work aims to reignite interest in leveraging expert features in remote sensing tasks and offers promising avenues for improved SAR image interpretation. Full article

► Show Figures

Figure 1

29 pages, 2991 KiB

Open AccessArticle

Embedding Moving Baseline RTK for High-Precision Spatiotemporal Synchronization in Virtual Coupling Applications

by Susu Huang, Baigen Cai, Debiao Lu, Yang Zhao, Miao Zhang and Linyu Shang

Remote Sens. 2025, 17(7), 1238; https://doi.org/10.3390/rs17071238 - 31 Mar 2025

Abstract

Achieving high-precision spatiotemporal synchronization is crucial for the implementation of virtual coupling (VC) in railway systems. This paper proposes a moving baseline real-time kinematic (MB-RTK) framework to enhance relative positioning accuracy and synchronization robustness between coupled trains. By leveraging global navigation satellite system [...] Read more.

Achieving high-precision spatiotemporal synchronization is crucial for the implementation of virtual coupling (VC) in railway systems. This paper proposes a moving baseline real-time kinematic (MB-RTK) framework to enhance relative positioning accuracy and synchronization robustness between coupled trains. By leveraging global navigation satellite system (GNSS) carrier-phase differential processing and dynamic baseline estimation, MB-RTK effectively mitigates positioning errors caused by GNSS signal degradation, multipath interference, and synchronization latency, ensuring stable and reliable inter-train coordination. The proposed framework was evaluated through comprehensive simulations and field experiments. The results demonstrate that MB-RTK achieves centimeter-level relative positioning accuracy under normal GNSS conditions, maintains tracking errors within 10 m, and typically keeps velocity synchronization deviations within ±0.5 km/h. Furthermore, the RTK status analysis reveals that NARROW_INT provides the highest stability, while continuous RTK corrections are essential to ensure seamless synchronization in dynamic environments. To further enhance synchronization performance, a decentralized distributed synchronization algorithm was introduced, reducing communication overhead and improving real-time responsiveness. The proposed approach exhibits strong resilience to GNSS disruptions, making it well-suited for high-density and autonomous train operations. Overall, this study highlights MB-RTK as a promising solution for VC applications, offering high accuracy, low latency, and strong adaptability in complex railway scenarios. Future research will focus on AI-driven dynamic corrections, integration with complementary localization methods, and large-scale deployment strategies to further optimize the system’s robustness and scalability. Full article

(This article belongs to the Special Issue Intelligent Perception and Robust Positioning Methods in GNSS-Denied Environments)

22 pages, 4563 KiB

Open AccessArticle

Attribute-Based Learning for Remote Sensing Image Captioning in Unseen Scenes

by Zhang Guo, Haomin Liu, Zihao Ren, Licheng Jiao, Shuiping Gou and Ruimin Li

Remote Sens. 2025, 17(7), 1237; https://doi.org/10.3390/rs17071237 - 31 Mar 2025

Abstract

Remote sensing image captioning (RSIC) aims to describe ground objects and scenes within remote sensing images in natural language form. As the complexity and diversity of scenes in remote sensing images increase, existing methods, although effective in specific tasks, are largely trained on [...] Read more.

Remote sensing image captioning (RSIC) aims to describe ground objects and scenes within remote sensing images in natural language form. As the complexity and diversity of scenes in remote sensing images increase, existing methods, although effective in specific tasks, are largely trained on particular scene images and corpora. This limits their ability to generate descriptions for scenes not encountered during training. Given the finite resources for data annotation and the expanding range of application scenarios, training data typically cover only a subset of common scenes, leaving many potential scene types unrepresented. Consequently, developing models capable of effectively handling unseen scenes with limited training data is imperative. This study introduces an innovative remote sensing image captioning model based on scene attribute learning—SALCap. The proposed model defines scene attributes and employs a specifically designed global object scene attribute extractor to capture these attributes. It then uses an attribute inference module to predict scene information through scene attributes, ensuring that this part of the scene’s information is reused in sentence generation through additional attribute loss. Experiments show that the method not only improves the accuracy of the description but also significantly enhances the model’s adaptability and generalizability relative to unseen scenes. This advancement expands the practical utility of remote sensing image captioning across diverse scenarios, particularly under the constraints of limited annotations. Full article

(This article belongs to the Special Issue Advancements of Vision-Language Models (VLMs) in Remote Sensing)

► Show Figures

Figure 1

30 pages, 19388 KiB

Open AccessArticle

An Explainable CatBoost Model for Crater Classification Based on Digital Elevation Model

by Minghao Zhu, Jialong Lai, Xiaoping Zhang, Yi Xu and Weidong He

Remote Sens. 2025, 17(7), 1236; https://doi.org/10.3390/rs17071236 - 31 Mar 2025

Abstract

The study of secondary craters on the Moon is vital for understanding lunar impact dynamics and surface evolution. However, this task is complicated by sample imbalance, with primary crater samples outnumbering those of secondary craters, and by the reliance on time-intensive manual methods [...] Read more.

The study of secondary craters on the Moon is vital for understanding lunar impact dynamics and surface evolution. However, this task is complicated by sample imbalance, with primary crater samples outnumbering those of secondary craters, and by the reliance on time-intensive manual methods or limited automated techniques. While many previous studies have focused on the manual or automated differentiation of secondary craters, few have addressed the interpretation of variables and models. In this study, we propose a machine-learning-based approach using the CatBoost algorithm to classify craters based on variables extracted from Digital Elevation Model (DEM) data. These variables include those from previous research as well as new ones introduced here, such as slope and density with Gaussian summation. Despite data imbalance and noise, the model achieves a classification accuracy of 0.8788, with a precision of 0.7922, a recall rate of 0.7412, and an F1 score of 0.7658 for secondary craters. To enhance interpretations, Shapley additive explanations (SHAP) and partial dependence plots (PDPs) are applied to evaluate variable importance and visualize the marginal effects of key variables, indicating the density variables playing a key role in crater classification. Full article

(This article belongs to the Special Issue High-Resolution Observations of Planetary Geological and Geomorphic Investigation (Second Edition))

► Show Figures

Figure 1

22 pages, 4131 KiB

Open AccessArticle

Improving AOD Algorithm Evaluation: A Spatial Matching Method for Minimizing Quality Control Bias

by Bailin Du, Bo Zhong, He Cai, Shanlong Wu, Yang Qiao, Xiaoya Wang, Aixia Yang, Junjun Wu, Qinhuo Liu, Jinxiong Jiang and Haizhen Zhang

Remote Sens. 2025, 17(7), 1235; https://doi.org/10.3390/rs17071235 - 31 Mar 2025

Abstract

Satellite-derived aerosol optical depth (AOD) products from MODIS and VIIRS sensors are vital for monitoring global aerosol distributions. However, inconsistencies in quality control algorithms and spatial resolution introduce errors that complicate validation processes and reduce the accuracy of satellite-to-ground comparisons. This study proposes [...] Read more.

Satellite-derived aerosol optical depth (AOD) products from MODIS and VIIRS sensors are vital for monitoring global aerosol distributions. However, inconsistencies in quality control algorithms and spatial resolution introduce errors that complicate validation processes and reduce the accuracy of satellite-to-ground comparisons. This study proposes the “optimal” spatial matching method to minimize these errors and enable a more accurate evaluation of retrieval algorithm performance. Using AERONET ground observations from 2012 to 2021, MODIS and VIIRS AOD products were systematically validated with three spatial matching methods—“direct”, “average”, and “optimal”. Results demonstrate that the “optimal” method consistently outperformed the other methods by selecting pixel values. The study highlights significant quality control disparities across AOD products and demonstrates that high-resolution products, with purer pixels, achieve superior accuracy under the “optimal” method. These insights provide valuable guidance for optimizing dataset applications and refining aerosol retrieval algorithms. Full article

► Show Figures

Figure 1

Journal Description

Remote Sensing

Latest Articles

Journal Menu

Journal Browser

Highly Accessed Articles

Latest Books

E-Mail Alert

News

Topics

Conferences

Special Issues

Topical Collections

Further Information

Guidelines

MDPI Initiatives

Follow MDPI