Search Results (6,639)

Terahertz (THz) waves offer unique advantages, including non-contact operation, high penetration capability, and high resolution, making them particularly well-suited for the non-destructive thickness measurement of film-structured materials. In reflective terahertz time-domain spectroscopy (THz-TDS), thickness measurement approaches are generally classified into three categories: optimization-based methods that rely on theoretical models, time-of-flight (ToF), and machine learning. Model-based optimization techniques require precise knowledge of the optical parameters and structural configuration of each layer; however, they often suffer from slow convergence and are prone to becoming trapped in local optima. In contrast, ToF-based methods determine thickness by calculating the time delay between echo pulses reflected from different interfaces, yet their applicability is limited when the film thickness is extremely small. Machine learning, especially deep learning, enables the establishment of a direct, data-driven mapping between THz waveforms (or their extracted features) and the target thickness. Such approaches offer rapid inference, strong robustness to noise, and good adaptability to thin or structurally complex films, although their accuracy remains dependent on the quality of training data and the generalization capability of the model. In this study, high-fidelity THz waveform data generated via finite-difference time-domain (FDTD) simulations are utilized to conduct a comparative investigation into the film thickness prediction performance of several representative machine learning algorithms, including Back Propagation (BP) neural networks, Support Vector Machines (SVM), Random Forests (RF), Extreme Learning Machines (ELM), K-Nearest Neighbors (KNN), and Partial Least Squares (PLS) regression. The results indicate that, in terms of prediction error, the overall ranking of algorithmic performance from best to worst is: PLS > RF > SVM > BP > ELM > KNN. These findings provide valuable guidance for the future application of machine learning-assisted THz-TDS in precise film thickness measurement. Full article

Change detection (CD) is a fundamental remote sensing task that identifies surface modifications from multi-temporal imagery of the same area, with applications in urban monitoring, agriculture, forest disturbance mapping, disaster assessment, and land cover analysis. The task is complicated by radiometric and atmospheric variability, co-registration errors, seasonal cycles, and sensor heterogeneity. Deep learning has progressively superseded traditional and classical machine learning approaches through hierarchical feature extraction and end-to-end optimization. Following the PRISMA 2020 guidelines, this systematic review examines 144 primary studies identified through a structured Scopus search complemented by the authors’ prior research and citation searching, spanning three paradigms: traditional approaches (algebraic operators, transformations, probabilistic frameworks), classical machine learning (support vector machines, random forests, object-based analysis), and deep learning architectures (fully convolutional, Siamese, attention-based, Transformer, state space, diffusion-based, and weakly supervised models). We provide background on problem formulation, benchmark datasets, and evaluation metrics, alongside a taxonomy organized by paradigm and supervision mode. A quantitative comparison on dominant benchmarks reveals the strengths and limitations of current methods. Open challenges include the absence of a universal benchmark protocol, the research-to-deployment gap, and the need for label-efficient learning. This review serves as a structured reference and outlines promising directions for the field. Full article

Accurate crop classification is essential for sustainable agriculture activities and food security studies. Recent advancements in remote sensing data acquisition and analysis techniques enable various solutions for cropland detection; however, reliable crop maps are still lacking in many heterogeneous semi-arid regions (e.g., Central Asia). Machine learning approaches address such challenges and distinguish different crop types using multiple datasets. The main aim of this study is to optimize crop classification outcomes by integrating multi-sensor datasets leveraging numerous vegetation indices through different machine learning models. Four datasets: Landsat-8 (DS-1), Sentinel-2 (DS-2), optical Sentinel-2 integrated with SAR Sentinel-1 (DS-3), and Sentinel-1 (DS-4) were used for the developed experiments. Five vegetation indices, NDVI, GNDVI, EVI, SAVI, and MSAVI, were derived using Sentinel-2 and Landsat-8 bands; in addition, NDRE was only obtained for Sentinel-2 exploiting the red edge band. Three input scenarios were considered for model training and image classification, featuring solely NDVI and its related bands; a set of vegetation indices and their associated bands for optical imagery; and VV, VH, and VV/VH ratio bands for SAR data. Five classifiers, Gradient Boosting Tree (GBT), Random Forest (RF), K-Nearest Neighbor (KNN), Classification and Regression Tree (CART), and Minimum Distance (MD), were employed to assess the machine learning quality for scene classification. Findings demonstrated that Sentinel-2 outperforms Landsat-8 images due to the higher spatial resolution and red edge bands. DS-3 consistently outperforms both DS-2 (optical-only) and DS-4 (SAR-only) across all classifiers, enhancing the overall accuracy up to 2.38% over the optical dataset, and up to 13.28% over the SAR data, demonstrating the added details on canopy spectral reflectance, structure and moisture content. Using multiple vegetation indices consistently improves performance over NDVI alone across DS-1, DS-2, and DS-3, with gains reaching up to 96.22% due to the complementary information captured by multi-index spectral sensitivity. The GBT and RF classifiers consistently achieved the highest classification performance, effectively combining multiple decision trees to capture complex nonlinear relationships and decision boundaries; meanwhile, the MD classifier exhibited the lowest accuracy due to its reliance solely on distances to class mean vectors. All in all, the presented approach offers a robust framework for crop classification supplemented with multiple data sources using different VI feature scenarios and variable machine learning tools for precise farming applications in semi-arid regions. Full article

High-precision estimation of forest aboveground biomass (AGB) is crucial for global carbon cycle monitoring and sustainable forest management. However, existing machine learning-based approaches often suffer from high-dimensional feature redundancy, uneven spatial distribution of training samples, and inefficient hyperparameter optimization, which collectively limit estimation accuracy and computational efficiency. To address these issues, this study proposes a synergistic feature-sample optimization framework (DFS) for high-precision forest AGB estimation. First, with the involvement of forestry experts, we constructed the Hunan and Hubei datasets covering typical subtropical forest types through multi-source remote sensing and ground plot sampling. Second, we propose the Dual-Criteria Adaptive Feature Selection (DCAFS) method, integrating ReliefF and mutual information criteria to adaptively select key features highly correlated with AGB, eliminating spectral redundancy while preserving biomass-sensitive information. Next, we introduce a Bidirectional Active Learning Sample Optimization mechanism, called BALSO, and in its forward step, plots with high uncertainty and representativeness are given priority, so samples with high AGB variability can be captured effectively; in the backward step, spatially redundant samples and feature-redundant samples are removed through density peak clustering, and by doing this, sample selection and spatial distribution are optimized at the same time, so plot balance gets improved. Finally, the framework brings in a parameter tuning structure based on Dream Optimization Algorithm, namely DOA, and through staged exploration together with local fine-tuning, DOA makes model hyperparameters and AGB data distribution characteristics align in an adaptive manner, which helps improve convergence efficiency and estimation stability. Input variables comprise Landsat 8 OLI spectral bands, GLCM texture features, vegetation indices, and Sentinel-1/2 data. On the Hunan dataset, the framework achieved an $R^2$ of 0.83 and an RMSE of 25.6 Mg·ha⁻¹; on the Hubei dataset, it achieved an $R^2$ of 0.86 and an RMSE of 26.8 Mg·ha⁻¹. The framework was further validated on an independent public dataset from Inner Mongolia. These results demonstrate that the DFS framework provides an effective and feasible approach for regional-scale forest AGB estimation and carbon monitoring. Full article

Rockfall hazards are frequently observed within mountainous road networks. Significant uncertainties regarding the optimal selection of evaluation units and spatial modeling scales are still being identified in this field. A comprehensive comparative framework for rockfall susceptibility mapping is presented in this study, using Wufeng County as the empirical study area. Five evaluation scenarios were constructed to systematically isolate the independent predictive contributions of the spatial domain, the mapping unit morphology, and the physics-informed engineering proxy. These scenarios included a whole-county macro-scale raster; three multi-scale road buffers with widths of 1 km, 2 km, and 3 km; and an object-oriented vector road evaluation unit (REU) framework. To parameterize localized engineering-induced risks, a physics-informed feature defined as the theoretical slope-cutting height ($H_{\mathrm{cut}}$) was structurally introduced into the vector-based assessment. Thirteen representative machine learning, deep learning, and statistical algorithms—including Random Forest, LightGBM, and TabNet—were systematically cross-examined under both unconstrained splits and strict Leave-One-Road-Corridor-Out Validation (LORCOV) protocols. The empirical multi-metric sensitivity analysis explicitly decouples the three structural effects. First, isolating the effect of the spatial domain reveals that restricting the validation extent from a broad countywide area to a narrow road corridor purges unperturbed background terrain noise, shifting the focus from easy negatives to geomorphological hard negatives. Second, evaluating the independent effect of the evaluation unit demonstrates that transitioning from continuous raster pixels to homogeneous vector REUs successfully resolves the terrain smoothing effect, precisely characterizing sharp geomechanical gradients adjacent to cut slopes. Third, isolating the effect of adding $H_{\mathrm{cut}}$ proves that this engineering indicator drives the primary descriptive gain, enabling tree-based ensembles to achieve a peak baseline AUC of 0.7763 and maintain a robust spatial validation AUC of 0.6129 under strict geographic block constraints, whereas legacy deep learning architectures exhibit an inductive bias mismatch on small-scale tabular records. Rather than asserting a single optimal paradigm, this coordinated feature–unit matching framework provides transport authorities with a highly calibrated, target-tiered decision matrix to optimize localized public works safety budgets and protect critical linear infrastructure assets. Full article

Mountainous road corridor landslides pose major dangers to infrastructure and transportation networks in areas with rugged terrain, fractured limestone lithology, and road-induced instability. This research presents a segment-based Random Forest (RF) model to evaluate landslide susceptibility along a mountainous road corridor spanning about 20 km in northeastern Iraq. Ten conditioning factors for landslide susceptibility were determined using Google Earth Engine (GEE) and GIS analysis. The model, validated using a 70/30 train/test split, achieved a mean cross-validation AUC of 0.783 ± 0.072 and an independent test AUC of 0.725 (accuracy = 0.735; Cohen’s Kappa = 0.401; recall = 0.750). To carry out independent multi-scale validation, the RF susceptibility maps were compared with a high-resolution SLAM LiDAR–AHP susceptibility approach within an overlapping ~2 km subsection, providing cross-scale validation of corridor-scale RF susceptibility predictions using a high-resolution susceptibility mapping framework. Comparison of the two approaches showed high spatial agreement, with 88.9% of the 18 overlapping segments exhibiting exact or one-class agreement between the two approaches. Full article

Predicting the mechanical properties of bolted connection nodes in prefabricated Concrete-Filled Steel Tube (CFST) column–steel beam assemblies remains challenging due to complex nonlinear relationships, high degrees of parameter coupling, and limited generalization capabilities of traditional empirical formulas. This study proposes a data-driven prediction model integrating an Improved Particle Swarm Optimization (IPSO) algorithm with a Deep Neural Network (DNN). Drawing upon 196 sets of experimental data on CFST column–steel beam nodes with Extended Hollo-Bolt (EHB) connections from the published literature, the model employs bolt diameter, steel tube wall thickness, concrete compressive strength, beam–column cross-sectional parameters, and connection configuration parameters as input variables, while designating ultimate moment capacity, initial stiffness, and joint ductility coefficient as prediction targets. A multi-layer DNN is constructed to capture the highly nonlinear mapping between structural parameters and mechanical responses. The IPSO algorithm, enhanced with adaptive inertia weight and Lévy flight perturbation, performs global optimization of the network weights and hyperparameters to improve convergence speed and prediction stability. Five-fold cross-validation is embedded within the IPSO fitness evaluation loop to guide hyperparameter selection, while dropout regularization and early stopping are applied during final training to mitigate overfitting; prediction performance is ultimately verified on an independent hold-out test set. Experimental results demonstrate that the proposed IPSO-DNN model outperforms a tuned shallow neural network (SNN), Support Vector Regression (SVR), and Random Forest (RF) models across the coefficient of determination ($R^2$), root mean square error (RMSE), and mean absolute error (MAE), effectively capturing the nonlinear mechanical characteristics of CFST nodes under complex loading conditions. Full article

To enhance the accuracy and real-time performance of anomaly detection for stay cables in cable-stayed bridges, especially those in water-rich environments, this paper proposes a stay cable anomaly identification method based on multi-point displacement measurements of the main girder. First, combined with finite element modeling, two anomaly identification indicators—“fitted displacement” and “effective displacement” of the main girder—are constructed, and their correlations with the location and severity of cable damage are clarified. Second, a gray level–displacement mapping model based on the Eulerian perspective is established, enabling non-contact, high-precision displacement measurement suitable for remote and distributed monitoring conditions. Subsequently, using a scaled cable-stayed bridge model in the laboratory, the feasibility and effectiveness of the method are verified through multi-condition experiments. Finally, the whale optimization algorithm (WOA)-optimized random forest model, enhancing the accuracy and stability of cable anomaly classification and identification. The results demonstrate that the proposed method offers significant advantages in accuracy, real-time performance, and intelligent recognition. It shows great potential for application in the health monitoring and safety management of water-related cable-stayed bridges, providing a new approach and technical support for intelligent safety diagnosis and reinforcement. Full article

Landslide susceptibility mapping (LSM) is fundamental to disaster prevention and spatial risk management in mountainous regions. However, conventional LSM approaches that rely mainly on static landslide influencing factors and empirical classification thresholds may have limited temporal relevance and interpretability. In response to these limitations, this study proposed an LSM framework constrained by interferometric synthetic aperture radar (InSAR)-derived deformation information. Wangmo County, Guizhou Province, China, was selected as the study area. Multi-source data, including small baseline subset InSAR (SBAS-InSAR) deformation results, optical remote sensing imagery, geo-environmental factors, and field investigation data, were used to construct and validate four machine learning models: logistic regression (LR), random forest (RF), support vector machine (SVM), and back-propagation neural network (BPNN). The validation results showed that the RF and BPNN models performed better than the LR and SVM models in terms of AUC, accuracy, precision, recall, and F1-score. Accordingly, an RF–BPNN combined model was constructed using an equal-weight averaging strategy. Shapley value analysis indicated that terrain- and rainfall-related factors made dominant contributions to landslide susceptibility prediction, a finding consistent with the landslide development characteristics in the study area. InSAR-derived deformation information was extracted from 31 Sentinel-1A images using SBAS-InSAR. A classification adjustment strategy based on kernel density estimation (KDE) and the Pearson correlation coefficient (PCC) was then used to identify the susceptibility classification scheme with relatively high spatial consistency with deformation activity during the observation period. The optimized classification scheme achieved a PCC value of 0.65, compared with 0.61 for the natural breaks classification, indicating a modest improvement in the spatial consistency between susceptibility zoning and deformation activity. The Xiangle and Namu landslides were used as representative cases to illustrate the adjustment effects of the deformation-constrained classification scheme. The proposed framework provides a practical approach for incorporating observation-period InSAR-derived deformation information into regional LSM and can support landslide monitoring and decision-making in complex terrains. Full article

In recent years, unmanned aerial vehicles (UAVs) have gradually become an important tool for forest fire monitoring due to their flexibility and wide-area observation capability. However, existing detection algorithms still struggle to achieve a balance between computational complexity and detection performance in diverse background conditions, and data for such scenarios remain limited. Therefore, this paper proposes a lightweight forest fire smoke detection model based on YOLOv10n. Specifically, we introduce the RepViTBlock module to enhance smoke feature extraction and improve detection accuracy with low computational cost. Meanwhile, we design a Lightweight Efficient Convolutional Detection head (LECD), which improves smoke target recognition and localization while reducing the number of parameters and computational overhead of the detection head. We also adopt the Minimum Point Distance Intersection over Union (MPDIoU) as the bounding-box regression loss function to improve the localization accuracy of smoke bounding-box regression. In addition, we construct a UAV-perspective Forest Fire Smoke (UFFS) dataset, which contains typical forest fire smoke, nearby thin smoke, distant small-scale smoke, and smoke under diverse background conditions. Experiments were conducted on both the UFFS dataset and the Wildfire Smoke V1 dataset. The experimental results show that, compared with the baseline model, the proposed model reduces the number of parameters by 36.7% and GFLOPs by 42.3% on the UFFS dataset, while improving mAP50 by 1.3% and mAP50-95 by 3.7%. In addition, recall increases by 3.2% and precision increases by 3.6%, indicating an improved trade-off between detection accuracy and model complexity. Full article

Accurate characterization of riverbed substrate from remote sensing imagery is essential for applications in fluvial geomorphology, habitat modeling, and river management. While recent advances in computer vision, particularly deep learning, have improved sediment mapping capabilities, their reliance on large annotated datasets and computational resources limits their broader applicability. This study presents a scalable workflow for categorical substrate classification using ultra-high-resolution aerial RGB orthoimagery in clear-water river environments. The approach integrates spectral information with statistical and structural texture descriptors derived from Gray-Level Co-occurrence Matrices (GLCM) and Local Binary Patterns (LBP), combined within a Random Forest classification framework. The methodology is structured as a semi-automated, five-stage workflow: (1) expert-based ground-truth substrate annotation; (2) feature set generation; (3) spatially aware model optimization; (4) full-domain classification; and (5) design-based validation for independent accuracy assessment. Model performance is evaluated using spatially aware cross-validation and design-based probability sampling to account for spatial autocorrelation and provide unbiased accuracy estimates. The method was applied in four geomorphologically distinct alpine river reaches, achieving design-based overall accuracy ranging from 70% to 88%. These results demonstrate that RGB-based approaches can achieve reliable reach-scale categorical substrate classification when combined with appropriate feature representation and rigorous validation strategies. However, limitations remain for visually similar or transitional substrate classes, particularly fine sediments such as sand and clay, which are difficult to distinguish consistently even during manual annotation. The workflow is implemented using open-source tools and is applicable to clear-water conditions where the riverbed remains optically visible. Full article

Background: Pelvic and cranial markers are often fragmented in mass disasters. The resilient talocrural joint serves as a valuable secondary site. This study presents a novel 12-landmark configuration on conventional 2D AP ankle radiographs as a rapid, cost-effective auxiliary sex estimation tool, bypassing complex 3D reconstructions. Methods: Radiographs of 200 contemporary Turkish adults (100 males, 100 females) were calibrated via PACS DICOM metadata. Twelve landmarks mapping distal tibial (T1–T6) and fibular (F7–F12) cortical topography were tracked to derive linear metrics. Models were evaluated using 5-fold cross-validation. Results: Optimized logistic regression achieved the highest independent test accuracy of 80.0%, outperforming random forests (72.5%) and support vector machines (70.0%), with an AUC of 0.870 (95% CI: 0.765–0.975). Tibial T2–T3 and T6–T1 were the strongest predictors (p < 0.001, Cohen’s d > 1.5). Although fibular F9–F10 lost significance after Bonferroni correction (p = 0.552), it maintained substantial multivariate weight, indicating spatial synergy. Conclusions: This interpretable, regularized logistic regression model trained on minimal linear ankle metrics shows that biological sex can be estimated with promising accuracy (80.0%) as an auxiliary tool capable of being seamlessly integrated into clinical PACS and active forensic workflows. Full article

Tropical forests sustain exceptional biodiversity and regulate global carbon and water cycles, yet degradation and incomplete recovery increasingly compromise these functions. We integrated bibliometric mapping with structured systematic synthesis to characterize research at the tropical-forest degradation–restoration interface, identify its most influential contributors and publications, and evaluate evidence on drivers, interventions, monitoring, and knowledge gaps. This study addresses three guiding scientific questions: (i) How has the field developed over time and across geographic space? (ii) Which authors, institutions, journals and publications have been most influential? (iii) What does the evidence indicate about degradation drivers, restoration strategies, monitoring approaches and knowledge gaps? Scopus and Web of Science were searched for peer-reviewed articles and reviews published from 1980 to 2025. After deduplication and PRISMA-based screening, 1075 publications were analyzed bibliometrically, and the 400 most-cited studies were coded against twenty predefined questions. Scientific output increased by 10.68% annually, with 483 publications appearing during 2020–2025. The corpus comprised 318 journals and 4470 authors. Forest Ecology and Management was the leading source (141 publications; 13.1%), while the twenty most productive journals accounted for 44.3% of the corpus. Brancalion P.H.S. was the most productive author (23 publications), followed by Chazdon R.L. and Tabarelli M. (20 each). Citation influence was concentrated: the twenty most-cited documents received 25.7% of all citations, led by Ribeiro M.C. (3339 citations), whereas Hua F. achieved the highest publication-year-normalized citation score among this group. Agricultural expansion, pasture establishment, and logging were the most frequently reported degradation pressures, but interactions among logging, fire, drought, and fragmentation were rarely quantified. Restoration evidence supported a context-dependent continuum from natural regeneration to assisted and active interventions, although planting and enrichment were more visible than direct passive–active comparisons. Carbon, biomass, plant diversity, and forest structure dominated outcome assessment, whereas fauna, ecological interactions, governance, and socioeconomic dimensions received less attention. Monitoring relied mainly on satellite imagery and field inventories, with limited evaluation of tool performance. Long-term trajectories and evidence from Africa, Southeast Asia, seasonally dry forests, and montane systems remained scarce. Overall, the field is rapidly consolidating but remains geographically and thematically uneven. Future research should quantify interacting degradation processes, evaluate multidimensional and long-term recovery, connect remote sensing with ecologically meaningful field indicators, and integrate governance and social conditions into restoration planning. The synthesis supports preventing further degradation and treating restoration as a context-specific complement, rather than a substitute, for protecting remaining native forests. Full article

Student dropout remains a persistent challenge in higher education, with significant academic, social and institutional implications. Although machine learning (ML) models for dropout prediction have proliferated, most select input variables based on data availability rather than theory and rarely report psychometric validation of the constructs used. This study addresses both gaps by developing and validating an early warning system at a Mexican university using exclusively pre-enrollment data from more than 46,000 student records from 2014 to 2025. Following the CRISP-DM methodology, eight theoretically grounded constructs, anchored in Tinto’s integration model, Bean’s attrition model, and Cabrera et al.’s persistence model, were operationalized from institutional intake questionnaires and assessed for internal consistency using Cronbach’s alpha prior to model training. Eight supervised ML algorithms were benchmarked across distance-based (Logistic Regression, SVM, AdaBoost, ANN) and tree-based (Random Forest, XGBoost, LightGBM, CatBoost) families. A tuned and isotonically calibrated Random Forest achieved the best overall performance (recall = 0.743, F1 = 0.524, ROC-AUC = 0.734, PR-AUC = 0.478) on a strictly held-out test set that was not used at any stage of model development. The 2023–2025 cohorts, whose dropout labels were not yet observable under the institutional definition, were scored prospectively to generate operational risk profiles. SHAP analysis identified high-school GPA, parental education, and household asset indices as dominant predictors, which mapped directly onto the three theoretical frameworks. These findings demonstrate that psychometrically grounded pre-enrollment data alone can support an operationally deployable dropout detection system, enabling proactive, evidence-based retention interventions from the first day of enrollment. Full article

Urban forests play a crucial role in sustaining habitable urban environments, and tree canopy height (TCH) is a key parameter for characterizing urban forest structure and ecological functions. However, accurate TCH retrieval using spaceborne lidar remains challenging in urban built-up areas because of the high spatial heterogeneity of urban land cover. This study proposes an urban-oriented method for ICESat-2 data processing and TCH extraction and evaluates it in Peoria, Illinois, USA. By integrating spectral constraints with photon spatial distribution patterns, the method further separates non-ground photons into vegetation photons and building photons. Individual tree crown polygons are then introduced as spatial constraints for TCH extraction, helping to overcome the limitations of regular grids in representing both the laser footprint scale and fine urban landscape detail. In addition, a non-exclusive photon-to-crown assignment strategy based on energy weighting allowed each photon to contribute to multiple crowns in proportion to the footprint energy intercepted by each crown. Among the 1391 crowns associated with vegetation photons, 286 met the thresholds for both total and high-weight photon counts and were retained for accuracy assessment. Validation against canopy heights derived from airborne laser scanning (ALS) reference data shows that the proposed method achieves an R² of 0.55 and an RMSE of 2.71 m. Further analysis indicates that retrieval accuracy is primarily controlled by beam strength: strong beams yield higher accuracy (R² ≈ 0.60), whereas daytime and nighttime observations differ only slightly. Overall, after targeted processing, ICESat-2 can provide a set of individual tree canopy height samples in urban built-up areas, which may support local calibration and serve as potential labels for subsequent regional canopy height mapping. Full article