Chang-Yong Yi

Accurate surface defect classification is a critical requirement in structural health monitoring and infrastructure inspection, where defects, including cracks, spalling, delamination and noncrack regions, often appear with low-contrast and complex background textures. Motivated by the need for a robust and discriminative framework that can enhance defect visibility and focus learning on damage-critical regions, this research proposes a novel damage-aware DenseNet-201 (DA-DenseNet-201) model for surface defect classification. As a critical novelty, a damage-aware adaptive contrast-limited adaptive histogram equalisation (DAC) filtering strategy is introduced as a preprocessing stage. The proposed DAC filter dynamically adjusts contrast enhancement parameters based on damage indicators, selectively amplifying crack edges and defect textures while preserving healthy surface regions and suppressing noise. Building on this method, enhanced images are processed using a pretrained DenseNet-201 backbone, retaining the benefits of dense feature propagation and efficient gradient flow. To strengthen the discriminative learning of DA-DenseNet-201 further, an attention refinement block is integrated into the network, combining channel attention to emphasise defect-relevant feature responses and spatial attention to localise damage regions accurately. In addition, a multiscale feature fusion mechanism aggregates feature maps from multiple dense blocks to capture fine-grained crack patterns, texture-level degradation and high-level semantic damage information. Extensive experiments conducted on surface defect datasets demonstrate its effectiveness, achieving a superior classification accuracy of 98.93%, along with notable improvements in sensitivity, specificity and the intersection over union compared with state-of-the-art models. These results confirm that the proposed DA-DenseNet-201 provides a reliable and high-performance solution for automated surface defect classification.

Accurate prediction of wind direction is a critical requirement for coastal safety management, renewable energy optimization, and weather-driven risk mitigation, particularly in highly dynamic atmospheric environments where statistical and deep learning models often struggle to capture nonlinear interactions and temporal dependencies. Existing approaches typically rely on raw or weakly processed meteorological inputs and treat directional information implicitly, which limits their ability to exploit the underlying physical structure of wind evolution. To address these challenges, this research designs a novel Physics Vector Driven (PVD) data pre-processing framework that explicitly encodes physically meaningful gradients and directional dynamics from multivariate meteorological observations, transforming raw measurements into sequence-aware vector representations suitable for deep time-series learning. Building on this foundation, a novel Directional Temporal Gradient Vector Network (DTGVectorNet) is proposed, which fuses a Directional Gradient Attention ResNet (DGResNet 1D CNN) for spatial-directional feature extraction with a Temporal Gradient LSTM (TGLSTM) designed to model the temporal evolution of wind vectors. The tight integration of Directional Gradient Attention (DGA) and Temporal Gradient (TG) memory enables the network to jointly learn instantaneous directional cues and their temporal propagation, significantly enhancing predictive fidelity. An experimental evaluation of the Busan wind datasets demonstrates that the proposed DTGVectorNet achieves a wind direction prediction accuracy of 99.12%, substantially outperforming conventional state-of-the-art baselines. These results confirm that physics-aware vector preprocessing combined with directional-temporal gradient fusion provides a powerful and generalizable paradigm for high-precision wind direction forecasting. To ensure reproducibility and facilitate further research, the complete dataset and implementation details of DTGVectorNet are publicly available through an open-access repository, Zenodo.

Automated recognition of celestial bodies from observational imagery is a cornerstone of autonomous space exploration. However, deploying deep learning models in space environments entails rigorous requirements not only for accuracy but also for reliability (calibration) and safety (anomaly rejection). Traditional Convolutional Neural Networks (CNNs) trained on small-scale astronomical datasets often suffer from overfitting and overconfidence on Out-of-Distribution (OOD) artifacts. In this work, we present a robust classification framework based on DINOv2, a Vision Transformer pre-trained via discriminative self-supervised learning. We curate a high-fidelity dataset of seven planetary classes sourced from NASA archives and propose a two-stage domain adaptation strategy to transfer large-scale foundation model features to this fine-grained task. Extensive experiments show that our method reaches 100% Top-1 accuracy on the canonical split, and remains highly stable under split variation, achieving 99.43% ± 0.85% Top-1 accuracy across R = 5 repeated stratified splits. More importantly, we address the critical issue of model trustworthiness. Through post hoc temperature scaling, our model achieves a state-of-the-art Expected Calibration Error (ECE) of 0.08%, representing a 36-fold improvement over ResNet50 (2.90%) and a 4.5-fold improvement over the EfficientNet-B3 baseline (0.36%). Furthermore, by integrating Energy-based OOD detection, the system effectively rejects non-planetary artifacts with an AUROC of 93.7%. Qualitative analysis using Grad-CAM reveals that self-supervised attention mechanisms naturally focus on intrinsic planetary features (e.g., surface textures and rings) while ignoring background noise, confirming the superior robustness of vision foundation models in astronomical vision tasks.

Urban green-space monitoring in dense cityscapes remains limited by accuracy–efficiency trade-offs and the absence of integrated, auditable area estimation. We introduce GreenViT, a Vision Transformer (ViT) based framework for precise segmentation and transparent quantification of urban green space. GreenViT couples a ViT-L/14 backbone with a lightweight single-path, progressive upsampling decoder (Green Head), preserving global context while recovering thin structures. Experiments were conducted on a manually annotated dataset of 20 high-resolution satellite images collected from Satellites.Pro, covering five land-cover classes (background, green space, building, road, and water). Using a 224 × 224 sliding window sampling scheme, the 20 images yield 62,650 training/validation patches. Under five-fold evaluation, it attains 0.9200 ± 0.0243 mIoU, 0.9580 ± 0.0135 Dice, and 0.9570 PA, and the calibrated estimator achieves 1.10% relative area error. Overall, GreenViT strikes a strong balance between accuracy and efficiency, making it particularly well-suited for thin or boundary-rich classes. It can be used to support planning evaluations, green-space statistics, urban renewal assessments, and ecological red-line verification, while providing reliable green-area metrics to support urban heat mitigation and pollution control efforts. This makes it highly suitable for decision-oriented long-term monitoring and management assessments.

Bioinspired morphing offers a powerful route to higher aerodynamic and hydrodynamic efficiency. Birds reposition feathers, bats extend compliant membrane wings, and fish modulate fin stiffness, tailoring lift, drag, and thrust in real time. To capture these advantages, engineers are developing airfoils, rotor blades, and hydrofoils that actively change shape, reducing drag, improving maneuverability, and harvesting energy from unsteady flows. This review surveys over 296 studies, with primary emphasis on literature published between 2015 and 2025, distilling four biological archetypes—avian wing morphing, bat-wing elasticity, fish-fin compliance, and tubercled marine flippers—and tracing their translation into morphing aircraft, ornithopters, rotorcraft, unmanned aerial vehicles, and tidal or wave-energy converters. We compare experimental demonstrations and numerical simulations, identify consensus performance gains (up to 30% increase in lift-to-drag ratio, 4 dB noise reduction, and 15% boost in propulsive or power-capture efficiency), and analyze materials, actuation, control strategies, certification, and durability as the main barriers to deployment. Advances in multifunctional composites, electroactive polymers, and model-based adaptive control have moved prototypes from laboratory proof-of-concept toward field testing. Continued collaboration among biology, materials science, control engineering, and fluid dynamics is essential to unlock robust, scalable morphing technologies that meet future efficiency and sustainability targets.

Fire doors are installed between compartments to prevent the spread of fire. During a fire, the temperature difference between the exposed and unexposed surfaces induces bending deformation of the door, thereby reducing its fire resistance performance. Excessive deformation may further compromise the structural integrity of the door. This study presents a thermo-mechanical model that idealizes the bending behavior of double swing fire doors based on the deflection equation of a simply supported beam subjected to a thermal gradient between the tensile and compressive sides. A criterion of deformation, quantifying the relationship between the meeting stile gap and the resulting maximum deflection, is introduced and compared with the predicted values. The validity of the proposed model was confirmed through fire resistance tests conducted on both insulated and non-insulated fire door specimens, demonstrating strong agreement with experimental results. Furthermore, by comparing the predicted deformation with the deformation criterion, the impact of increasing gap sizes on the service life of fire doors on their fire resistance performance was evaluated. Based on this analysis, appropriate gap size limits for different door specifications are proposed to ensure reliable fire performance.

In this study, we delve into a novel approach by employing a sensor-based pattern recognition model to address the automation of construction equipment activity analysis. The model integrates time of flight (ToF) sensors with deep convolutional neural networks (DCNNs) to accurately classify the operational activities of construction equipment, focusing on piston movements. The research utilized a one-twelfth-scale excavator model, processing the displacement ratios of its pistons into a unified dataset for analysis. Methodologically, the study outlines the setup of the sensor modules and their integration with a controller, emphasizing the precision in capturing equipment dynamics. The DCNN model, characterized by its four-layered convolutional blocks, was meticulously tuned within the MATLAB environment, demonstrating the model’s learning capabilities through hyperparameter optimization. An analysis of 2070 samples representing six distinct excavator activities yielded an impressive average precision of 95.51% and a recall of 95.31%, with an overall model accuracy of 95.19%. When compared against other vision-based and accelerometer-based methods, the proposed model showcases enhanced performance and reliability under controlled experimental conditions. This substantiates its potential for practical application in real-world construction scenarios, marking a significant advancement in the field of construction equipment monitoring.

Point cloud data‐based edge line extraction is an important task for accurate geometrical inspection of precast concrete (PC) elements in the construction industry. Although a few edge extraction algorithms have been developed so far based on point cloud data, little attention has been paid on which edge extraction algorithm performs the best in terms of edge estimation accuracy. To tackle the research gap, this study aims to evaluate currently available edge extraction algorithms in order to determine optimal algorithm for precise geometrical inspection of PC elements. To do this, simulated scan points are first generated and used for algorithm performance analysis using a geometrical model and a measurement noise modeling that determine the coordinates of simulated scan points. For validation of the simulation approach, comparison tests with experimental data are performed and the results show that the simulation approach has a high similarity of more than 90% compared to experimental data in terms of the number of scan points, scan pattern, and scan density, proving the effectiveness of the simulation‐based evaluation method. In addition, it shows that a least square regression (LSR) algorithm provides the best performance with an edge extraction accuracy of 1.56 and 2.71 mm for simulated and experimental scan points, respectively. The contributions of this study are (1) development of the geometrical model and noise modeling based on actual scan data and (2) validation of simulated‐based evaluation method on the lab‐scale PC slabs.