Search Results (2,410)

Road defect detection is essential for timely road damage repair and traffic safety assurance. However, existing object detection algorithms suffer from insufficient accuracy in detecting small road surface defects and are prone to missed detections and false alarms under complex lighting and background conditions. To address these challenges, this study proposes SDC-YOLOv8, an improved YOLOv8-based algorithm for road defect detection that employs attention-enhanced feature learning and adaptive feature reconstruction. The model incorporates three key innovations: (1) an SPPF-LSKA module that integrates Fast Spatial Pyramid Pooling with Large Separable Kernel Attention to enhance multi-scale feature representation and irregular defect modeling capabilities; (2) DySample dynamic upsampling that replaces conventional interpolation methods for adaptive feature reconstruction with reduced computational cost; and (3) a Coordinate Attention module strategically inserted to improve spatial localization accuracy under complex conditions. Comprehensive experiments on a public pothole dataset demonstrate that SDC-YOLOv8 achieves 78.0% mAP@0.5, 81.0% Precision, and 70.7% Recall while maintaining real-time performance at 85 FPS. Compared to the baseline YOLOv8n model, the proposed method improves mAP@0.5 by 2.0 percentage points, Precision by 3.3 percentage points, and Recall by 1.8 percentage points, yielding an F1 score of 75.5%. These results demonstrate that SDC-YOLOv8 effectively enhances small-target detection accuracy while preserving real-time processing capability, offering a practical and efficient solution for intelligent road defect detection applications. Full article

Energy-efficient driving is essential for reducing the environmental impacts of road transport, especially for electric passenger vehicles. This research aims to build a data-driven behavioral analysis and energy-consumption evaluation model. The model relies on sensor data from the vehicle’s on-board communication network, primarily the CAN (Controller Area Network) bus. We analyze patterns of key powertrain and battery parameters—such as current, voltage, state of charge (SoC), and power—in relation to driver inputs, such as the accelerator pedal position. In the first stage, we review the literature with a focus on machine learning and clustering methods used in behavioral and energy analysis. We also examine the role of on-board telemetry systems. Next, we develop a controlled measurement architecture. It defines reference consumption maps from dynamometer data across operating points and environmental variables, including SoC, temperature, and load. The longer-term goal is a multidimensional behavioral map and profiling framework that can predict energy efficiency from real-time driver inputs. This work lays the foundation for a future system with adaptive, feedback-based driver support. Such a system can promote intelligent, sustainable, and behavior-oriented mobility solutions. Full article

Traditional routing algorithms optimizing for distance or travel time are inadequate for electric vehicles (EVs), which require energy-aware planning considering battery constraints and charging infrastructure. This work presents an energy-optimal routing system for EVs that integrates personalized consumption modeling with real-time environmental data. The system employs a Long Short-Term Memory (LSTM) neural network to predict State-of-Charge (SoC) consumption from real-world driving data, learning directly from spatiotemporal features including velocity, temperature, road inclination, and traveled distance. Unlike physics-based models requiring difficult-to-obtain parameters, this approach captures nonlinear dependencies and temporal patterns in energy consumption. The routing framework integrates static map data, dynamic traffic conditions, weather information, and charging station locations into a weighted graph representation. Edge costs reflect predicted SoC drops, while node penalties account for traffic congestion and charging opportunities. An enhanced A* algorithm finds optimal routes minimizing energy consumption. Experimental validation on a Nissan Leaf shows that the proposed end-to-end SoC estimator significantly outperforms traditional approaches. The model achieves an RMSE of 36.83 and an $R^2$ of 0.9374, corresponding to a 59.91% reduction in error compared to physics-based formulas. Real-world testing on various routes further confirms its accuracy, with a Mean Absolute Error in the total route SoC estimation of 2%, improving upon the 3.5% observed for commercial solutions. Full article

Gabal (G.) Suwayqat El-Arsha contains two distinct phases of granitoids: I-type granodiorite and A-type monzogranite. Both of them experienced intense fractional crystallization that affected plagioclase, alkali feldspar, quartz, and, to a lesser degree, ferromagnesian minerals. EnMAP hyperspectral data were used to discriminate between the different granitoid types through spectral analysis, using various techniques, including the Sequential Maximum Angle Convex Cone (SMACC) method. Granodiorite has high SiO₂ (68.21–71.44 wt%), Al₂O₃ (14.29–14.92 wt%), Fe₂O₃ (1.99–3.32 wt%), and CaO (2.34–3.87 wt%), whereas monzogranite has even higher SiO₂ (73.58–75.87 wt%) and K₂O (4.28–4.88 wt%). Both granodiorite and monzogranite exhibit calc-alkaline, peraluminous to metaluminous, and medium- to high-K characteristics, with attendant enrichment of light REE and LILE and depletion of heavy REE and HFSE. A negative Eu anomaly may indicate early plagioclase fractionation, especially in the monzogranite. The I-type granodiorite is likely derived from a high-K, mafic protolith that partially melted during lithospheric delamination, leading to severe fractional crystallization in the upper crust in a post-collisional environment. In contrast, the monzogranite exhibits A-type characteristics and was likely emplaced in an anorogenic setting. Both granites were affected by several episodes of hydrothermal alteration, resulting in silicification, kaolinitization, sericitization, and chloritization. The intrusions studied here exhibit key similarities with those in the Wadi El-Hima area, including tectonic setting, petrogenetic type, Neoproterozoic age (Stage I collisional: ca. 650–620 Ma; Stage II post-collisional: ca. 630–590 Ma), and mineralogical assemblages (notably two-mica granites). These correlations suggest that both suites form part of a regionally extensive batholith composed of I- and A-type granites, stretching from north of the Marsa Alam Road (Umm Salatit–Homrit Waggat) southward to at least Wadi El-Hima. Full article

High-resolution remote sensing images are gradually playing an important role in land cover mapping, urban planning, and environmental monitoring tasks. However, current segmentation approaches frequently encounter challenges such as loss of detail and blurred boundaries when processing high-resolution remote sensing imagery, owing to their complex backgrounds and dense semantic content. In response to the aforementioned limitations, this study introduces HMA-UNet, a novel segmentation network built upon the UNet framework and enhanced through a hybrid attention strategy. The architecture’s innovation centers on a composite attention block, where a lightweight split fusion attention (LSFA) mechanism and a lightweight channel-spatial attention (LCSA) mechanism are synergistically integrated within a residual learning structure to replace the stacked convolutional structure in UNet, which can improve the utilization of important shallow features and eliminate redundant information interference. Comprehensive experiments on the WHDLD dataset and the DeepGlobe road extraction dataset show that our proposed method achieves effective segmentation in remote sensing images by fully utilizing shallow features and eliminating redundant information interference. The quantitative evaluation results demonstrate the performance of the proposed method across two benchmark datasets. On the WHDLD dataset, the model attains a mean accuracy, IoU, precision, and recall of 72.40%, 60.71%, 75.46%, and 72.41%, respectively. Correspondingly, on the DeepGlobe road extraction dataset, it achieves a mean accuracy of 57.87%, an mIoU of 49.82%, a mean precision of 78.18%, and a mean recall of 57.87%. Full article

To address the structural differences between closed-track and open-road testing in terms of scenario coverage, risk controllability, and validation consistency, this study proposes a scenario-driven combined testing method for automated driving systems. The proposed approach constructs a multi-dimensional scenario space based on functional decomposition and jointly quantifies scenario complexity and hazard level from the perspectives of information heterogeneity and interaction-induced risks. Based on these two-dimensional scenario attributes, a fuzzy inference mechanism is developed to dynamically allocate validation resources across different testing environments. To further improve rule-base generalization and mapping stability, an enhanced genetic algorithm integrating simulated annealing and K-means clustering is introduced to optimize the rule structures in an evolutionary manner. Experimental results demonstrate that, compared with traditional testing methods and single-mechanism optimization strategies, the proposed approach achieves a more consistent and interpretable mapping between scenarios and testing proportions in high-complexity urban traffic scenarios. While ensuring test adequacy, the testing economy is significantly improved, with an overall average improvement exceeding 20%. In addition, stable resource allocation performance is observed across multiple scenarios with different levels of complexity and risk, confirming the scalability and applicability of the proposed method for multi-scenario automated driving system testing. Full article

The Iraqi coastline in the northern Persian Gulf is highly vulnerable to the impacts of future sea level rise. This study introduces a novel approach in the Arc Geographic Information System (ArcGIS) for inundation risk of the 58 km Iraqi coast of the northern Persian Gulf through a combination of multi-data sources, machine-learning predictions, and hydrological connectivity by Landsat. The Prophet/Neural Prophet time-series framework was used to extrapolate future sea level rise with 11 satellite altimetry missions that span 1993–2023. The coastline was obtained by using the Landsat-8 Operational Land Imager (OLI) imagery based on the Normalised Difference Water Index (NDWI), and topography was obtained by using the ALOS World 3D 30 m DEM. Global Land Use and Land Cover (LULC) projections (2020–2100) and population projections (2020–2100) were used as future inundation values. Two scenarios were compared, one based on an altimeter-based projection of sea level rise (SLR) and the other based on the National Aeronautics and Space Administration (NASA) high-emission scenario, Representative Concentration Pathway 8.5 (RCP8.5). It is found that, by the IPCC AR6 end-of-century projection horizon (relative to 1995–2014), 154,000 people under the altimeter case and 181,000 people under RCP8.5 will have a risk of being inundated. The highest flooded area is the barren area (25,523–46,489 hectares), then the urban land (5303–5743 hectares), and finally the cropland land (434–561 hectares). Critical infrastructure includes 275–406 km of road, 71–99 km of electricity lines, and 73–82 km of pipelines. The study provides the first hydrologically verified Digital Elevation Model (DEM)-refined inundation maps of Iraq that offer a baseline, in the form of a comprehensive and quantitative base, to the coastal adaptation and climate resilience planning. Full article

Forest fires pose escalating threats to ecological security and public safety in Guangdong Province. This study presents a novel machine learning framework for fire occurrence prediction by synergistically integrating multi-source geospatial data. Utilizing Moderate-resolution Imaging Spectroradiometer (MODIS) active fire detections from 2014 to 2023, we quantified historical fire patterns and incorporated four categories of predisposing factors: meteorological variables, topographic attributes, vegetation characteristics, and anthropogenic activities. Spatiotemporal clustering dynamics were characterized via kernel density estimation and spatial autocorrelation analysis. An XGBoost classifier, hyperparameter-optimized through the Sparrow Search Algorithm (SSA), achieved a predictive accuracy of 90.4%, with performance evaluated through precision, recall, and $F_1$-score. Risk zoning maps generated from predicted probabilities were validated against independent fire records from 2019 to 2024. Results reveal pronounced spatial heterogeneity, with high-risk zones concentrated in northern and western mountainous areas, constituting 29% of the provincial territory. Critical driving factors include slope gradient, proximity to roads and rivers, temperature, population density, and elevation. This robust predictive framework furnishes a scientific foundation for spatially-explicit fire prevention strategies and optimized resource allocation in key high-risk jurisdictions, notably Qingyuan, Shaoguan, Zhanjiang, and Zhaoqing. Full article

Traffic flow anomalies represent significant deviations from normal traffic behavior and disrupt the smooth operation of transportation systems. These may appear as unusually high or low traffic volumes compared to historical trends. Unexpectedly high volume can lead to congestion exceeding usual capacity, while unusually low volume might indicate incidents like road closures, or malfunctioning traffic signals. Identifying and understanding both types of anomalies is crucial for effective traffic management. This paper presents a clustering based approach for enhanced characterization of anamolies in traffic flows. Anomalies in traffic patterns are determined using three anomaly detection techniques: Elliptic Envelope, Isolation Forest, and Local Outlier Factor. These anomalies were newly detected in this work on the Montréal dataset after preprocessing, rather than directly reused from earlier studies. These methods were applied to a dataset that had been pre-processed using windowing techniques with different configuration settings to enhance the detection process. Then, to leverage the detected anomalies, we utilized clustering algorithms, specifically k-means and hierarchical clustering, to segment these anomalies. Each clustering algorithm was used to determine the optimal number of clusters. Subsequently, we characterized these clusters through detailed visualization and mapped them according to their unique characteristics. This approach not only identifies traffic anomalies effectively but also provides a comprehensive understanding of their spatial and temporal distributions, which is crucial for traffic management and urban planning. Full article

Road marking lines can be extracted from aerial images using semantic segmentation (SS) models; however, in this work, a conditional generative adversarial network, RoadMark-cGAN, is proposed for direct extraction of these representations with image-to-image translation techniques. The generator features residual and attention blocks added in a functional bottleneck, while the discriminator features a modified PatchGAN, with an optimized encoder and an attention block added. The proposed model is improved in three versions (v2 to v4), in which dynamic dropout techniques and a novel “Morphological Boundary-Sensitive Class-Balanced” (MBSCB) loss are progressively added to better handle the high class imbalance present in the data. All models were trained on a novel “RoadMarking-binary” dataset (29,405 RGB orthoimage tiles of 256 × 256 pixels and their corresponding ground truth masks) to learn the distribution of road marking lines found on pavement. The metrical evaluation on the test set containing 2045 unseen images showed that the best proposed model achieved average improvements of 45.2% and 1.7% in the Intersection-over-Union (IoU) score for the positive, underrepresented class when compared to the best Pix2Pix and SS models, respectively, trained for the same task. Finally, a qualitative, visual comparison was conducted to assess the quality of the road marking predictions of the best models and their mapping performance. Full article

The M_JMA 7.6 (Mw 7.5) Noto Peninsula Earthquake of 1 January 2024 in Japan triggered widespread slope failures across northern Noto region, but their spatial controls and susceptibility patterns remain poorly quantified. Most previous studies have focused mainly on fault rupture, ground deformation, and tsunami impacts, leaving a clear gap in machine learning based assessment of earthquake-induced slope failures. This study integrates 2323 mapped landslides with eleven conditioning factors to develop the first data-driven susceptibility framework for the 2024 event. Spatial analysis shows that 75% of the landslides are smaller than 3220 m² and nearly half occurred within about 23 km of the epicenter, reflecting concentrated ground shaking beyond the rupture zone. Terrain variables such as slope (mean 31.8°), southwest-facing aspects, and elevations of 100–300 m influenced the failure patterns, along with peak ground acceleration values of 0.8–1.1 g and proximity to roads and rivers. Six supervised machine learning models were trained, with Random Forest and Gradient Boosting achieving the highest accuracies (AUC = 0.95 and 0.94, respectively). Explainable AI using SHapley Additive exPlanations (SHAP) identified slope, epicentral distance, and peak ground acceleration as the dominant predictors. The resulting susceptibility maps align well with observed failures and provide an interpretable foundation for post-earthquake hazard assessment and regional risk reduction. Further work should integrate post-seismic rainfall, multi-temporal inventories, and InSAR deformation to support dynamic hazard assessment and improved early warning. Full article

Driver fatigue detection based on electroencephalography (EEG) signals has gained increasing attention for enhancing road safety. However, existing deep learning models often treat EEG data as generic time-series inputs, neglecting the inherent hierarchical and spatial–temporal structure of brain activity, which limits their interpretability and generalization. To address this, we propose a novel Residual Polynomial Network (RPN) that explicitly models the positive and negative activation patterns in EEG signals through a polarity-aware architecture. The RPN integrates polarity decomposition, residual learning, and hierarchical feature fusion to capture discriminative neurophysiological dynamics while maintaining model transparency. Extensive experiments are conducted on a real-world driving fatigue dataset using a subject-wise 10-fold cross-validation protocol. Results show that the proposed RPN achieves an average classification accuracy of 97.65%, outperforming conventional machine learning and deep learning baselines including SVM, KNN, DT, and LSTM. Ablation studies confirm the effectiveness of each component, and Sankey diagram analysis provides interpretable insights into feature-to-class mappings. This work not only advances the state of the art in EEG-based fatigue detection but also offers a more transparent and physiologically plausible deep learning framework for brain signal analysis. Full article

In recent years, deep learning-based image classification has made significant progress, especially in safety-critical perception fields such as intelligent vehicles. Factors such as vibrations caused by NVH (noise, vibration, and harshness), sensor noise, and road surface roughness pose challenges to robustness and real-time deployment. The Transformer architecture has become a fundamental component of high-performance models. However, in complex visual environments, shifted window attention mechanisms exhibit inherent limitations: although computationally efficient, local window constraints impede cross-region semantic integration, while deep feature processing obstructs robust representation learning. To address these challenges, we propose DAR-Swin (Dual-Attention Revamped Swin Transformer), enhancing the framework through two complementary attention mechanisms. First, Scalable Self-Attention universally substitutes the standard Window-based Multi-head Self-Attention via sub-quadratic complexity operators. These operators decouple spatial positions from feature associations, enabling position-adaptive receptive fields for comprehensive contextual modeling. Second, Latent Proxy Attention integrated before the classification head adopts a learnable spatial proxy to integrate global semantic information into a fixed-size representation, while preserving relational semantics and achieving linear computational complexity through efficient proxy interactions. Extensive experiments demonstrate significant improvements over Swin Transformer Base, achieving 87.3% top-1 accuracy on CIFAR-100 (+1.5% absolute improvement) and 57.0% mAP on COCO2017 (+1.3% absolute improvement). These characteristics are particularly important for the active and passive safety features of intelligent vehicles. Full article

Reliable traffic-signal information delivery is critical for safe navigation through signalized intersections, particularly for low-cost autonomous vehicles that rely on Vehicle-to-Network (V2N) communication rather than on-board HD maps or expensive perception sensors. Ensuring this selective delivery requires accurate infrastructure-side map-matching, which becomes challenging when vehicles operate with only Standard Definition (SD) maps and noisy GNSS measurements. This study comparatively evaluates five infrastructure-side map-matching algorithms under varying GNSS errors and road-network densities using real trajectories from Jeju Island with controlled Gaussian perturbations. The framework includes geometric matching, Extended Kalman Filtering (EKF), route-constrained filtering, grid-based spatial indexing, and a hybrid route–EKF fallback mechanism, executed in real time on a cloud-hosted Kafka pipeline. The hybrid route–EKF algorithm exhibited consistently high and stable link-matching accuracy (0.99308–0.96546 across GPS error groups; 0.9887–0.9777 across density groups) together with strong signal-matching accuracy (0.99394–0.96950; 0.9865–0.9790). Route-constrained and Kalman-based approaches also performed well, while heading-based matching showed clear limitations. These results indicate that infrastructure-side map-matching provides a scalable foundation for cloud-assisted traffic-signal information services and supports the feasibility of delivering reliable traffic-signal information to low-cost autonomous platforms. Full article

A well-designed nighttime lighting environment not only enhances pedestrian comfort and urban vitality but also serves as a crucial factor in creating safe and livable urban spaces. However, existing studies on pedestrian safety at night remain relatively limited both domestically and internationally, and most rely primarily on ground-based measurements, making large-scale spatial analyses difficult to achieve. This study integrates night-time remote sensing, ground measurements and perception evaluations to analyze the light environments of three public space types—roads, business districts and squares—in Dalian, China. A light environment safety perception model and corresponding map are constructed to support optimization of lighting design in urban nightscapes. The main contributions are as follows: (1) subjective and objective research conducted on the night light environment safety perception of urban public space in Dalian; (2) fitting models are developed for each space type to relate measured illuminance to perceived safety, yielding recommended ground illuminance ranges: roads (4.02–10.10 lx), business districts (5.05–38.3 lx), and squares (6.46–12.52 lx); (3) models linking nighttime radiation data to measured illuminance are established, enabling the generation of an illuminance inversion map for Dalian. Based on this, safety classification maps for roads, business districts, and squares are produced. These are integrated with the residential area safety map to form a comprehensive safety classification map of Dalian’s urban area. Full article