Search Results (2,075)

Accurate quantification of forest biomass in the karst mountainous region of Southwest China is critical for regional carbon sink accounting. However, remote sensing-based inversion of forest biomass in this region is subject to the dual constraints of scale effects and spatial heterogeneity. Taking two typical karst landforms—plateau karst and peak-cluster depression karst—as case studies, this study developed an inversion framework that integrates multi-scale synergy (plot—small watershed—region) with geomorphological zoning modeling. Specifically, the high-precision forest aboveground biomass (AGB), retrieved by integrating high-resolution satellite imagery (2 m) of small watersheds with field plot data, was used as the scale-conversion bridge. The dominant class variability-weighted method was applied to upscale the spatial resolution from 2 m to 30 m. Subsequently, Landsat-8 OLI imagery, land use/land cover data, and topographic factors were integrated to construct landform-specific neural network models, namely BPANN-GY for plateau karst and BPANN-FC for peak-cluster depression karst, with validation RMSEs of 11.02% and 12.39%, respectively. The results showed that the mean forest AGB in the plateau karst region was 110.66 t·ha⁻¹ in 2013 and 111.46 t·ha⁻¹ in 2024; in the peak-cluster depression karst region, the mean forest AGB was 123.98 t·ha⁻¹ in 2014 and 125.57 t·ha⁻¹ in 2024. Compared with the baseline years (2013/2014), total forest AGB in both regions increased significantly by 2024, up to 11.43% in the plateau karst region and 6.43% in the peak-cluster depression karst region. Spatially, the most marked increase in AGB occurred in the mid-to-high slope zones, while the differences in forest AGB among slope grades gradually narrowed, indicating progressively enhanced forest structural integrity and functional stability under effective land management. This study methodologically validates the scientific soundness and feasibility of using high-precision AGB retrieved from high-resolution satellite imagery as a scale-conversion bridge. The established AGB inversion system provides a methodological framework and data foundation for carbon sink accounting, ecological restoration, and land management in the karst region of Southwest China, and offers a transferable approach for forest AGB inversion in other highly heterogeneous landscapes. Full article

Accurate and up-to-date information is essential for environmental monitoring, particularly in regions characterized by complex topography and heterogeneous landscapes. This study presents a multisource remote sensing–based approach for forest vegetation classification on La Palma Island (Canary Islands, Spain), which was further used to illustrate its potential for monitoring the temporal dynamics of different forest habitat types. Very high-resolution multispectral data from the WorldView-2/3 satellites were used, complemented by multispectral and LiDAR data acquired by an unmanned aerial vehicle (UAV). Four target forest vegetation types were mapped within a six-class classification scheme that also included “Other vegetation” and “Soil/Others” as non-target/background classes. The performance of ten supervised classification algorithms was evaluated, including Minimum Distance, Mahalanobis Distance, Parallelepiped, Spectral Angle Mapper, Maximum Likelihood, Naïve Bayes, K-Nearest Neighbors, Random Forest, Support Vector Machine, and the transformer-based deep learning model SegFormer. The results indicate that Random Forest achieved the highest overall accuracy, while Support Vector Machine and SegFormer also showed competitive performance, particularly when spectral information was integrated with vegetation indices and topographic variables. The study provides practical evidence on the selection of input data and classifiers for detailed forest vegetation mapping in a large and topographically complex island. Full article

Remote sensing image super-resolution (RSISR) provides an effective means of improving spatial detail for Earth observation and satellite image interpretation. However, existing methods often rely on increasingly complex network designs with deeper hierarchies and expanded channel capacities to pursue higher performance, resulting in heavy models with high computational cost, which restricts their deployment on resource-constrained platforms. To address this challenge, we propose a novel reparameterized feature enhancement network (RepFEN) for lightweight and accurate RSISR tasks. Specifically, a multi-scale reparameterized module (MRepM) is designed to capture multi-scale spatial information and enhance texture representation. Furthermore, a partial-channel gated attention module (PCGAM) is introduced to selectively enhance discriminative features along the channel dimension, effectively improving fine-grained detail restoration. By integrating structural reparameterization and multi-scale lightweight modules, the proposed method achieves a better balance between reconstruction accuracy and inference efficiency. Extensive experiments on both remote sensing and natural image super-resolution benchmarks demonstrate that our method achieves superior performance compared to existing state-of-the-art methods, while maintaining minimal computational overhead, showing significant potential for real-world applications. Full article

In Deep Learning Remote Sensing, data quantity is rarely the limiting factor. A single high-resolution satellite image can yield thousands of training patches. What determines model performance, yet remains largely overlooked, is the quality of those patches. To date, the choice of sampling method has rarely been treated as a methodological decision. Conventional approaches, namely sliding-window and random sampling, introduce two compounding data-quality problems: severe class imbalance caused by the overproduction of background-only patches and negative learning arising from incomplete annotations, where unlabeled objects are implicitly treated as negative examples during training. To address these limitations at the data construction stage, we propose object-centric patch sampling, a model-independent strategy that anchors each training patch to the geometric centroid of an annotated object. This design ensures that every object-anchored patch contains at least one target instance and substantially reduces exposure to unlabeled regions that generate false-negative supervision signals; only a small, deliberately controlled proportion of background-only patches is retained to preserve contextual variety without reinstating background dominance. The method is evaluated on three heterogeneous remote sensing datasets spanning satellite (Sentinel-2, 10 m), aerial (NAIP, 1 m), and UAV (0.25 m) imagery, covering cotton field segmentation, rural building extraction, and water body delineation, respectively. Using a U-Net architecture under identical training conditions, the proposed approach achieves IoU scores of 0.929, 0.896, and 0.912 on the three datasets, respectively, outperforming sliding-window sampling by up to 19.6 percentage points in IoU and consistently delivering higher F1-scores across all experimental configurations. Evaluation under DeepLabV3+ gives a mean IoU of 0.9504 and a mean F1-score of 0.9772 in a multi-class segmentation task, indicating that the gains are not specific to a single architecture. Unlike model-level solutions such as focal loss or class reweighting, the proposed method improves training data quality at its source and integrates seamlessly into any deep learning pipeline without architectural modifications. Full article

Object detection in remote sensing imagery remains challenging due to vast scale variations, complex backgrounds, and the prevalence of small, densely packed targets. Existing CNN-based detectors are limited by restricted receptive fields, while Transformer-based methods incur prohibitive computational overhead for high-resolution inputs. In this paper, we propose SFSMamba-DETR, a detection framework that integrates state space models with Dual-Scale Window Attention for efficient and accurate remote sensing object detection. Specifically, we design a Selective Feature Scanning (SFS) module that uses the Mamba-based 2D Selective Scan mechanism to model long-range spatial dependencies with linear computational complexity. To capture both fine-grained local patterns and broader contextual cues simultaneously, we introduce a Dual-Scale Window Attention (DSWA) mechanism that operates at two complementary window scales with multi-kernel convolution bridging. These modules are orchestrated within a Cross-scale Feature Aggregation Module (CFAM) that performs hierarchical multi-scale fusion in a hybrid encoder. Extensive experiments on three primary benchmarks (MAR20, UCAS-AOD, and the Jilin-1 Satellite Aircraft Detection Dataset), together with supplementary results on DOTA and DIOR, demonstrate that SFSMamba-DETR achieves strong detection accuracy while maintaining competitive inference speed. Full article

The Greater Cape Floristic Region (GCFR) in South Africa is a fire-prone biodiversity hotspot where high species richness, structural complexity, and small plant sizes (0.0001–4 m²) pose substantial challenges for remote sensing-based biodiversity assessment. Spectral similarity among species and the mismatch between plant size and sensor pixel dimensions limit the capacity of current and forthcoming spaceborne systems to resolve individual species and accurately detect plot-level diversity changes. We therefore developed a physics-based simulation framework that couples fynbos trait measurements with radiative transfer modeling in the DIRSIG (Digital Imaging and Remote Sensing Image Generation) environment towards quantifying information loss across spectral and spatial scales and to define theoretical limits for biodiversity monitoring. We constructed a three-dimensional virtual scene of post-fire fynbos communities in Grootbos Private Nature Reserve, integrating high-resolution imagery, terrestrial laser scanning (TLS), and structure-from-motion (SfM)-derived point clouds. Field measurements of mean diameter and percent cover were used to scale vegetation models and constrain species abundance. We distributed plant instances using a blue noise sampling algorithm, guided by density maps derived from unmanned aerial system (UAS) imagery. Species-specific optical properties were parameterized using field-measured reflectance data and the PROSPECT radiative transfer model, while terrain structure was derived from SfM-based digital terrain models. The integrated scene was used to simulate multispectral (DJI Mavic 3 MSI), hyperspectral (AVIRIS-NG), and light detection and ranging (LiDAR) observations. Agreement between simulated outputs were evaluated against corresponding field-acquired datasets using spectral signatures and vegetation indices. This framework enables systematic assessment of sensor specification effects on spectral biodiversity metrics and provides a pathway for evaluating theoretical limits of species discrimination across airborne and satellite platforms. Full article

Environmental monitoring of freshwater ecosystems faces significant challenges in detecting illicit wastewater discharges, which often remain unrecognized due to their intermittent nature and limited spatial footprint. This study presents a novel integrated strategy combining satellite remote sensing, Unmanned Aerial Vehicle (UAV)-based proximal sensing, and in situ measurements to enhance pollution detection in vulnerable aquatic environments. The methodology was applied to Lake Avernus (Italy), a volcanic lake historically affected by eutrophication and toxic cyanobacterial blooms. Landsat 8–9 thermal analysis revealed no detectable anomalies, reflecting the limitations of its coarse spatial resolution. Sentinel-2 multispectral imagery was then analyzed through spectral indices, band ratios, and reflectance signatures, revealing localized variations in surface reflectance and spatial heterogeneity in water optical properties. These satellite-derived anomalies guided targeted high-resolution UAV surveys. UAV-based thermal imaging revealed an elevated-temperature zone along the adjacent shoreline. In situ field screening flagged a candidate chemical anomaly at this location. The hierarchical framework demonstrates that satellite screening effectively identifies areas of concern, while UAV thermal imaging enables high-resolution localization of features invisible to satellite sensors, and in situ measurements provide essential ground-truth validation. This replicable, low-cost methodology offers a powerful tool for early warning, surveillance, and sustainable management of sensitive freshwater ecosystems. Full article

Remote sensing of clouds and precipitation is undergoing a significant transition from descriptive observations toward process-oriented diagnoses. This transition has not progressed linearly, but has gradually taken place alongside the rapidly developing multi-source observation capabilities over recent decades. In terms of clouds and precipitation observational platforms, satellites provide continuous global-scale monitoring, airborne platforms complement high-resolution sampling of key processes, and ground-based observations offer long-term vertical structure evolution, which form an integrated space–air–ground observation system. In terms of clouds and precipitation retrieval algorithms, active–passive combination remote sensing significantly improves the ability to retrieve macro- and microphysical characteristics and structures, and machine learning methods further expand parameter estimation capabilities in complex scenarios. Nevertheless, key bottlenecks still persist in retrieval non-uniqueness, sensor trade-offs, cross-platform calibration, and validation over oceans, mountains, and polar regions. Based on the above background, this paper provides a systematic review of recent progress in clouds and precipitation physics remote sensing, focusing on the development of multi-platform collaborative observations, the evolution of microphysical parameter retrieval methods, and the improvements in remote sensing characterization of cloud and precipitation formation mechanisms. It further points out that future development will increasingly rely on improved uncertainty quantification, incorporation of physical constraints into retrieval frameworks, and the establishment of standardized multi-source datasets. Full article

High-resolution remote sensing stereo matching is challenging due to heavy computation and the difficulty of handling textureless areas, repetitive structures, and occlusions. To tackle these issues, we design BDNet (Balancing Dual-path Network)—a stereo matching network that seeks a reasonable trade-off between accuracy and efficiency for remote sensing applications. In the feature extraction stage, BDNet adopts progressive dilation with rates 5, 4, and 3, together with a decoupled multi-scale reduction (DMSR) module, which reduces multi-scale feature channels from 320 to 32. We also introduce a strip attention module to make the network more sensitive to horizontal and vertical structures commonly seen in urban scenes. For cost volume construction, the number of correlation groups is lowered from 40 to 8, in line with the compact 32-channel feature representation. For cost aggregation, a dual-path parallel hourglass architecture is designed, which preserves fine details through a high-resolution path while capturing global context through a low-resolution path. An attention-guided fusion module adaptively integrates features from both paths, improving accuracy in challenging regions such as textureless areas and disparity discontinuities. Experiments on the US3D and WHU-Stereo datasets demonstrate that BDNet achieves the best accuracy among the selected baseline methods on US3D, with D1 errors of 16.05% on Jacksonville and 11.84% on Omaha. It requires only 1.43 M parameters and 125.43 G FLOPs, achieving a favorable balance between accuracy and efficiency. Zero-shot generalization experiments on Omaha and WHU-Stereo further suggest the model’s potential for cross-domain adaptation to different satellite sensors and urban scenes. Full article

Ocean sound speed profile (SSP) is a key parameter for underwater acoustic detection, remote sensing, and seafloor geodetic positioning, and its temporal prediction is essential for improving acoustic positioning accuracy. Conventional direct measurements are inefficient and spatially sparse, while statistical and acoustic inversion methods fail to capture the strong nonlinear evolution of the sound speed field. Among existing time series models, LSTM, a recurrent network for time series forecasting, lacks an explicit receptive field. In contrast, the original TCN, a temporal convolutional network with dilated convolutions, poorly captures local fine structures and relies heavily on empirical tuning. To overcome these limitations, we propose an improved TCN-based SSP prediction method and apply it to seafloor geodetic positioning. The approach first constructs a sound speed increment field via first-order time differencing to remove global trends and highlight local variations. It then employs Optuna (version 4.9.0), a Bayesian sampling-based automatic optimization framework, to automatically tune key TCN parameters within a predefined search space, reducing reliance on manual tuning. The predicted high-resolution sound speed time series is finally used for ray tracing positioning to enhance seafloor geodetic accuracy. Experiments on the GLORYS12V1 reanalysis dataset show that LSTM and the original TCN achieve root mean square error (RMSE) and mean absolute error (MAE) values of 0.414 and 0.299 m/s, as well as 0.360 and 0.258 m/s, respectively, whereas our improved TCN reduces these to 0.205 and 0.131 m/s, substantially outperforming both baselines. In simulated Global Navigation Satellite System–Acoustics (GNSS-A) seafloor positioning, the 3D positioning RMSE drops to about 0.075 m, with improved stability. The proposed method offers an effective solution for accurate SSP time series forecasting and high-precision seafloor geodesy. Full article

Remote sensing is widely recognized as a key technology across a wide range of technical and scientific domains, especially in agriculture. Although satellite data have long supported crop monitoring, their limitations in spatial resolution, revisit frequency and cloud coverage have often constrained their applications. High-resolution satellites, available from the beginning of the 2000s, have improved performance, particularly in the field of precision agriculture, but they remain expensive and inflexible. Unmanned Aerial Vehicles perform better in precision agriculture, offering flexibility and high levels of detail; however, their limited operational areas and short endurance flight times constrain their effectiveness. In this evolving landscape, High Altitude Pseudo Satellites (HAPSs), particularly high-altitude balloons, are emerging as a promising new technology that could fill the gaps between satellite and drone remote sensing. These platforms provide large area coverage with high-resolution imagery and long endurance flights at low operational expenses and ease of deployment. This study investigates the operational characteristics, strengths, and geometric limitations of data acquired by the CubeHAPS^® platform, a high-altitude pseudo-satellite system, as a prerequisite for its application in precision agriculture. Focusing on experimental campaigns conducted in northern Italy in summer 2024 and 2025, the research characterizes platform stability, image block consistency, and photogrammetric quality through internal metrics. The results demonstrate measurable improvements between the two campaigns, attributed to the introduction of a stabilization system in 2025 and establishing the conditions under which the platform can support reliable photogrammetric reconstruction. Full article

Accurate extraction of road networks from high-resolution remote sensing imagery is a fundamental task underpinning autonomous-driving navigation, urban spatial planning, and the dynamic updating of geographic information databases. Although existing road extraction methods attain outstanding pixel-level segmentation accuracy and topological integrity, most follow an accuracy-first design paradigm that relies on heavyweight backbones and increasingly complex decoders, incurring a parameter volume and storage overhead that constitute the principal bottleneck for deploying them on resource-constrained edge platforms such as unmanned aerial vehicles, mobile terminals, and onboard satellite processors. Conversely, models that pursue extreme lightweighting often fail to preserve the thin, continuous, linear structure of roads, tending to produce topological breaks in the extracted road networks. To bridge the performance gap between segmentation accuracy and model size, we propose LOA-Net, a lightweight orientation-aware road extraction network. LOA-Net introduces a Road-Aligned Deformable Convolution (RA-DCN) that adaptively aligns the sampling region with the road geometry and explicitly supervises the predicted road orientation, thereby accurately capturing road connectivity while substantially reducing the parameter count. Experiments on the CHN6-CUG and DeepGlobe benchmarks show that LOA-Net surpasses representative state-of-the-art methods on both IoU and F1, while achieving the lowest parameter count of all compared models and a computational complexity comparable to its peers, striking an excellent trade-off between segmentation performance and a mobile-friendly footprint that makes it well suited for road extraction from remote sensing imagery in resource-constrained scenarios. Full article

Effective control and management of the encroaching and invasive Vachellia karroo (V. karroo) in grassland and savanna biomes depends on accurate information about its spatial distribution, making remote sensing approaches essential for mapping the extent of affected areas. Although Sentinel-2 satellite data are widely used for land use and land cover applications, they often lack the spatial details required to distinguish woody species like V. karroo. The fusion of Sentinel-2 data with high-resolution UAV imagery offers a promising approach to enhance spectral information for species-level discrimination. This study evaluated UAV, Sentinel-2, and fused UAV–Sentinel-2 imagery for discrimination of V. karroo in grassland and savanna biomes of the Eastern Cape, South Africa. Field data and imagery were collected in October 2022 and classified using Random Forest (RF) and Support Vector Machine (SVM) algorithms to distinguish V. karroo. The findings showed that V. karroo was more prevalent in the savanna biome. SVM marginally outperformed RF in classifying V. karroo in the grassland biome, achieving overall accuracies ranging from 68.9% to 97.4%, compared to 57.8% to 97.4% for RF. Among the datasets, the fused UAV–Sentinel-2 images yielded the highest classification accuracy, with an overall accuracy of 97.4% and a kappa coefficient of 0.96. The UAV images also demonstrated high classification accuracy, with an overall accuracy of 91.67% and a kappa coefficient of 0.77, confirming its value for fine-scale mapping and reference data support. In contrast, the Sentinel-2 images produced lower classification accuracy, with an overall accuracy of 84.6% and a kappa coefficient of 0.75, mainly due to their coarser spatial resolution. Classification was more challenging in the savanna site, where mixed vegetation structure increased confusion between V. karroo and grass. These findings show that fused UAV–Sentinel-2 images can improve species-level discrimination, while UAV and Sentinel-2 data remain complementary for fine-scale mapping and broader monitoring of bush encroachment in grassland and savanna ecosystems. Full article

Mangroves provide critical ecosystem services, including coastal protection, biodiversity conservation, carbon sequestration, and environmental purification. Detailed mangrove mapping is therefore essential for effective conservation, restoration, and management. With the proliferation of remotely sensed imagery across diverse spatial resolutions, comparative studies on mapping efficiency are vital for optimizing long-term, large-scale monitoring strategies. This study utilized multi-source satellite imagery, which includes Landsat-8 (15 m and 30 m), Sentinel-2 (10 m), ZiYuan-3 (ZY-3), and GaoFen-1 (GF-1) (2 m), to map mangroves in the Beibu Gulf by integrating spectral, texture, and topographic features. The optimal feature combination was determined experimentally. We comprehensively compared the identification accuracy, identification results, and area estimates derived from three distinct methods: pixel-based Random Forest (RF), object-oriented RF, and a U-Net + ResNet-34 deep learning model. Key findings include the following: (1) Feature importance varied by resolution; terrain features significantly improved accuracy for medium-resolution images (e.g., Landsat-8), increasing the F1-score by 1.91%, while texture features were critical for high-resolution images (e.g., ZY-3 and GF-1), improving the F1-score by 2.99%. (2) Deep learning achieved the highest accuracy using Sentinel-2 combined with terrain features, yielding an F1-score of 96.45%. (3) Higher-resolution imagery enabled the detection of smaller mangrove patches; deep learning produced the fewest internal gaps, whereas pixel-based RF suffered from severe “salt-and-pepper” noise. (4) While object-oriented RF using Sentinel-2 imagery produced results most consistent with the validation set (recall = 94.38%), notable area discrepancies (9.7%) persisted, primarily due to tidal dynamics. We conclude that deep learning models, particularly when integrating Sentinel-2 data with terrain features, provide superior accuracy for mangrove mapping. Nevertheless, tidal dynamics continue to pose a significant challenge for computer-aided identification approaches. Full article

Remote-sensing object detection commonly requires repeated training on high-resolution aerial and satellite imagery, where targets may be small, densely distributed, and surrounded by extensive background. For remote-sensing detection tasks that require shorter model-training cycles, reducing training time without sacrificing detection accuracy is important. The Anti-Forgetting Sampling Strategy (AFSS) reduces training time by avoiding repeated processing of learned images, but its fixed warm-up may start sampling before the detector is mature and thereby reduce accuracy. We propose Maturity-Gated AFSS (MG-AFSS), a detector-maturity activation controller for AFSS. The method accumulates validation mean average precision at an intersection-over-union threshold of 0.50 (mAP50), fits a cumulative saturating curve online, and enables AFSS only when the estimated maturity indicates trustworthy image states. After activation, it reuses the original AFSS sampling rule. We evaluate MG-AFSS with matched lightweight detectors on five public remote-sensing datasets: NWPU VHR for the main three-seed study and additional settings; UCAS-AOD for horizontal-box detection; HRSC2016 and ShipRSImageNet for oriented bounding-box (OBB) boundary cases; and DOTAv1 for a larger-scale OBB block. In the three-seed, 80-epoch, from-scratch NWPU VHR experiment, MG-AFSS improves over AFSS by +0.0264 mAP50 and +0.0319 mAP50–95, where mAP50–95 denotes mean average precision averaged over intersection-over-union thresholds from 0.50 to 0.95. It keeps mAP50–95 essentially equal to standard full-dataset training (+0.0003), with a training-time ratio of 0.867 relative to standard training. In a pretrained NWPU setting, the training-time ratio is 0.557 relative to matched standard training, with a higher checkpoint mAP50. In the DOTAv1 block, MG-AFSS has a training-time ratio of 0.759, lower than AFSS at 0.777; its checkpoint-best differences from standard training are −0.0095 mAP50 and −0.0149 mAP50–95, both smaller in magnitude than those of AFSS. Overall, MG-AFSS yields a conditional accuracy–training-time trade-off: it reduces training time when the maturity condition is satisfied and, as observed in the two smaller OBB boundary runs, retains full-dataset training otherwise. Full article