Search Results (302)

Maritime unmanned aerial vehicle (UAV) ship detection faces challenges including variations in ship pose and appearance under multiple viewpoints, occlusion and confusion in dense scenes, complex backgrounds, and the scarcity of ship datasets from UAV tilted perspectives. To overcome these obstacles, this study introduces a high-quality dataset named Marship-OBB9, comprising 11,268 drone-captured images and 18,632 instances spanning nine typical ship categories. The dataset systematically reflects the characteristics of maritime scenes under diverse scales, viewpoints, and environmental conditions. Based upon this dataset, we propose a novel detection network named YOLO11-UAVShip. First, an oriented bounding box detection mechanism is incorporated to precisely fit ship contours and reduce background interference. Second, a newly designed CK_DCNv4 module, integrating deformable convolution v4 (DCNv4) and a C3k2 backbone structure, is developed to enhance geometric feature extraction under aerial oblique view. Additionally, for ships with large aspect ratios, SGKLD effectively addresses the localization challenges in dense environments, achieving robust position regression. Comprehensive experimental evaluation demonstrates that the proposed method yields a 2.1% improvement in mAP@0.5 and a 2.3% increase in recall relative to baseline models on the Marship-OBB9 dataset. While maintaining real-time inference speed, our approach greatly enhances detection accuracy and robustness. This work provides a practical and deployable solution for intelligent ship detection in UAV imagery. Full article

Traditional orebody modeling methods struggle to efficiently integrate new geological data. Therefore, we propose a novel framework for dynamically updating 3D geological models by directly interpolating geological logging data. The core innovation lies in the innovative interpolation of raw interpreted cross polylines into an implicit scalar field representation without intermediate explicit surface extraction or manual remodeling. To obtain reliable vectorized polylines, we developed image recognition and digitization techniques that are based on the pattern recognition of geological sketches. Moreover, different from existing implicit techniques, we present an improved approach to interpolate complex cross polylines that are dynamically based on the improved principal component analysis. The method allows specifying a priori constraints to adjust the erroneous estimated normal to improve the reliability of the normal estimation results of cross-contour polylines. The a priori information can be combined into the normal estimation algorithm to update the normals of the corresponding adjacent contour polylines in the process of normal estimation at the intersection points and in the process of normal propagation. By leveraging the radial basis functions-based spatial interpolators, the method continuously assimilates incremental geological observations into the interpolation constraints to update the implicit model. Case studies demonstrate a reduction in the modeling cycle time compared to conventional explicit methods while maintaining geologically coherent boundaries. The framework significantly enhances decision agility in resource estimation and mine planning workflows by bridging geological interpretation and dynamic model iteration. Full article

Droplet-based microlens fabrication using Ultra Violet (UV) curable polymers demands the precise measurement of three-dimensional geometries, especially for non-axisymmetric shapes influenced by electric field deformation. In this work, we present a polar coordinate-based contour fitting method combined with Hermite interpolation to reconstruct 3D droplet geometries from two orthogonal shadowgraphy images. The image segmentation process integrates superpixel clustering with active contours to extract the droplet boundary, which is then approximated using a spline-based polar fitting approach. The two resulting contours are merged using a polar Hermite interpolation algorithm, enabling the reconstruction of freeform droplet shapes. We validate the method against both synthetic Computer-Aided Design (CAD) data and precision-machined reference objects, achieving volume deviations below 1% for axisymmetric shapes and approximately 3.5% for non-axisymmetric cases. The influence of focus, calibration, and alignment errors is quantitatively assessed through Monte Carlo simulations and empirical tests. Finally, the method is applied to real electrically deformed droplets, with volume deviations remaining within the experimental uncertainty range. This demonstrates the method’s robustness and suitability for metrology tasks involving complex droplet geometries. Full article

Broken or loose strands in high-voltage transmission conductors constitute critical defects that jeopardize grid reliability. Unmanned aerial vehicle (UAV) inspection has become indispensable for their timely discovery; however, conventional detectors falter in the face of cluttered backgrounds and the conductors’ diminutive pixel footprint, yielding sub-optimal accuracy and throughput. To overcome these limitations, we present PowerStrand-YOLO—an enhanced YOLOv8 derivative tailored for UAV imagery. The method is trained on a purpose-built dataset and integrates three technical contributions. (1) A C2f_DCNv4 module is introduced to strengthen multi-scale feature extraction. (2) An EMA attention mechanism is embedded to suppress background interference and emphasize defect-relevant cues. (3) The original loss function is superseded by Shape-IoU, compelling the network to attend closely to the geometric contours and spatial layout of strand anomalies. Extensive experiments demonstrate 95.4% precision, 96.2% recall, and 250 FPS. Relative to the baseline YOLOv8, PowerStrand-YOLO improves precision by 3% and recall by 6.8% while accelerating inference. Moreover, it also demonstrates competitive performance on the VisDrone2019 dataset. These results establish the improved framework as a more accurate and efficient solution for UAV-based inspection of power transmission lines. Full article

The accurate segmentation of pigmented skin lesions is a critical prerequisite for reliable melanoma detection, yet approximately 30% of lesions exhibit fuzzy or poorly defined borders. This ambiguity makes the definition of a single contour unreliable and limits the effectiveness of computer-assisted diagnosis (CAD) systems. While clinical assessment based on the ABCDE criteria (asymmetry, border, color, diameter, and evolution), dermoscopic imaging, and scoring systems remains the standard, these methods are inherently subjective and vary with clinician experience. We address this challenge by reframing segmentation into three distinct regions: background, border, and lesion core. These regions are delineated using superpixels generated via the Simple Linear Iterative Clustering (SLIC) algorithm, which provides meaningful structural units for analysis. Our contributions are fourfold: (1) redefining lesion borders as regions, rather than sharp lines; (2) generating superpixel-level embeddings with a transformer-based autoencoder; (3) incorporating these embeddings as features for superpixel classification; and (4) integrating neighborhood information to construct enhanced feature vectors. Unlike pixel-level algorithms that often overlook boundary context, our pipeline fuses global class information with local spatial relationships, significantly improving precision and recall in challenging border regions. An evaluation on the HAM10000 melanoma dataset demonstrates that our superpixel–RAG–transformer (region adjacency graph) pipeline achieves exceptional performance (100% F1 score, accuracy, and precision) in classifying background, border, and lesion core superpixels. By transforming raw dermoscopic images into region-based structured representations, the proposed method generates more informative inputs for downstream deep learning models. This strategy not only advances melanoma analysis but also provides a generalizable framework for other medical image segmentation and classification tasks. Full article

We consider a two-step numerical approach for solving parabolic initial boundary value problems in 3D simply connected smooth regions. The method uses the Laplace transform in time, reducing the problem to a set of independent stationary boundary value problems for the Helmholtz equation with complex parameters. The inverse Laplace transform is computed using a sinc quadrature along a suitably chosen contour in the complex plane. We show that due to a symmetry of the quadrature nodes, the number of stationary problems can be decreased by almost a factor of two. The influence of the integration contour parameters on the approximation error is also researched. Stationary problems are numerically solved using a boundary integral equation approach applying the Nyström method, based on the quadratures for smooth surface integrals. Numerical experiments support the expectations. Full article

Precisely identifying and delineating food regions automatically from images, a task known as food image segmentation, is crucial for enabling applications in food science such as automated dietary logging, accurate nutritional analysis, and food safety monitoring. However, accurately segmenting food images, particularly delineating food edges with precision, remains challenging due to the wide variety and diverse forms of food items, frequent inter-food occlusion, and ambiguous boundaries between food and backgrounds or containers. To overcome these challenges, we proposed a novel method called the Multi-view Edge Attention Network (MVEANet), which focuses on enhancing the fine-grained segmentation of food edges. The core idea behind this method is to integrate information obtained from observing food from different perspectives to achieve a more comprehensive understanding of its shape and specifically to strengthen the processing capability for food contour details. Rigorous testing on two large public food image datasets, FoodSeg103 and UEC-FoodPIX Complete, demonstrates that MVEANet surpasses existing state-of-the-art methods in segmentation accuracy, performing exceptionally well in depicting clear and precise food boundaries. This work provides the field of food science with a more accurate and reliable tool for automated food image segmentation, offering strong technical support for the development of more intelligent dietary assessment, nutritional research, and health management systems. Full article

The structural safety monitoring of industrial heritage is of great significance for global urban renewal and the preservation of cultural heritage. However, traditional tilt monitoring methods suffer from limited accuracy, low efficiency, poor global perception, and a lack of intelligence, making them inadequate for meeting the tilt monitoring requirements of super-high-rise industrial heritage chimneys. To address these issues, this study proposes a tilt monitoring method for super-high-rise industrial heritage chimneys based on LiDAR point clouds. Firstly, LiDAR point cloud data were acquired using a ground-based LiDAR measurement system. This system captures high-density point clouds and precise spatial attitude data, synchronizes multi-source timestamps, and transmits data remotely in real time via 5G, where a data preprocessing program generates valid high-precision point cloud data. Secondly, multiple cross-section slicing segmentation strategies are designed, and an automated tilt monitoring algorithm framework with adaptive slicing and collaborative optimization is constructed. This algorithm framework can adaptively extract slice contours and fit the central axes. By integrating adaptive slicing, residual feedback adjustment, and dynamic weight updating mechanisms, the intelligent extraction of the unit direction vector of the central axis is enabled. Finally, the unit direction vector is operated with the x- and z-axes through vector calculations to obtain the tilt-azimuth, tilt-angle, verticality, and verticality deviation of the central axis, followed by an accuracy evaluation. On-site experimental validation was conducted on a super-high-rise industrial heritage chimney. The results show that, compared with the results from the traditional method, the relative errors of the tilt angle, verticality, and verticality deviation of the industrial heritage chimney obtained by the proposed method are only 9.45%, while the relative error of the corresponding tilt-azimuth is only 0.004%. The proposed method enables high-precision, non-contact, and globally perceptive tilt monitoring of super-high-rise industrial heritage chimneys, providing a feasible technical approach for structural safety assessment and preservation. Full article

Spilled hydrocarbons released from oil pipeline accidents can result in long-term environmental contamination and significant damage to habitats. In this regard, evaluating actions in response to vulnerability scenarios is fundamental to emergency management and groundwater integrity. To this end, understanding the trajectories and their influence on the various parameters and characteristics of the contaminant’s fate through accurate numerical simulations can aid in developing a rapid remediation strategy. This paper develops a numerical model using the CactusHydro code, which is based on a high-resolution shock-capturing (HRSC) conservative method that accurately follows sharp discontinuities and temporal dynamics for a three-phase fluid flow. We analyze nine different emergency scenarios that represent the breaking of a diesel oil onshore pipeline in a porous medium. These scenarios encompass conditions such as dry season rupture, rainfall-induced saturation, and varying pipeline failure pressures. The influence of the spilled oil pressure and water saturation in the unsaturated zone is analyzed by following the saturation contour profiles of the three-phase fluid flow. We follow with the high-accuracy formation of shock fronts of the advective part of the migration. Additionally, the mass distribution of the expelled contaminant along the porous medium during the emergency is analyzed and quantified for the various scenarios. The results obtained indicate that the aquifer contamination strongly depends on the pressure outflow in the vertical flow. For a fixed pressure value, as water saturation increases, the mass of contaminant decreases, while the contamination speed increases, allowing the contaminant to reach extended areas. This study suggests that, even for LNAPLs, the distribution of leaked oil depends strongly on the spill pressure. If the pressure reaches 20 atm at the time of pipeline failure, then contamination may extend as deep as two meters below the water table. Additionally, different seasonal conditions can influence the spread of contaminants. This insight could directly inform guidelines and remediation measures for spill accidents. The CactusHydro code is a valuable tool for such applications. Full article

In low-light optical imaging, the scarcity of incident photons and the inherent limitations of imaging sensors lead to challenges such as low signal-to-noise ratio, limited dynamic range, and degraded contrast, severely compromising image quality and optical information integrity. To address these challenges, we propose a novel low-light image enhancement technique, LightenResDiff, which combines a residual diffusion model with the discrete wavelet transform. The core innovation of LightenResDiff lies in it accurately restoring the low-frequency components of an image through the residual diffusion model, effectively capturing and reconstructing its fundamental structure, contours, and global features. Additionally, the dual cross-coefficients recovery module (DCRM) is designed to process high-frequency components, enhancing fine details and local contrast. Moreover, the perturbation compensation module (PCM) mitigates noise sources specific to low-light optical environments, such as dark current noise and readout noise, significantly improving overall image fidelity. Experimental results across four widely-used benchmark datasets demonstrate that LightenResDiff outperforms existing methods both qualitatively and quantitatively, surpassing the current state-of-the-art techniques. Full article

Rolling contact fatigue (RCF)-induced cracks in steel rails exhibit a fish-scale-shaped cluster distribution, and generally form in a layered, overlapping manner. Eddy-current-pulsed thermography (ECPT) has been applied in RCF detection by taking advantage of electromagnetic–thermal execution; however, one still faces challenges in identifying and quantifying such layered, overlapping defects. This paper proposes a swept-frequency eddy-current-pulsed thermal tomography (ECPTT) detection method to quantitatively characterize multilayer crack depth and inclination angle in an artificial rail sample. In particular, stimulating frequency modulation is used to guide the induced eddy current and heat to varying depths, and this is combined with principal component analysis (PCA) to identify multilayer defects. Moreover, a thermal signal reconstruction (TSR) algorithm is introduced. TSR features are extracted for analyzing the burial depth and inclination angle of multilayer defects. The results demonstrate that the third principal component (PC3), extracted via PCA, enables layer-count discrimination in multilayer defects. Integrated with gradient magnitude analysis of the second principal component (PC2) under swept-frequency excitation, defect contour localization error can be controlled within 0.5 mm. Building on layer discrimination, multi-frequency thermal response analysis further reveals variations in PC1’s variance contribution, differentiating inclination angles of 10° and 20°, whereas comparative heating- and cooling-rate magnitudes distinguish burial depths of 0.5 mm and 1.0 mm. The research verifies that the ECPTT system can accurately detect the layer number, inclination angle, and depth of buried RCF defects, substantially enhancing the accuracy of defect contour reconstruction. Full article

The present article provides an abstract overview of the issue of optimal blank searching for integral parts utilized in complex engineering projects, including those pertaining to the fabrication of machine, ship, and aircraft components. The manufacturing process for these components is intricate and necessitates meticulous precision and strict adherence to the design model. Conventional blank calculation techniques are marred by substantial inaccuracies. The present research proposes and verifies an effective method based on the reverse solution of a mathematical problem. The focal point of this study is the aerodynamic curvature of aluminum alloys belonging to the Al–Mg–Mn family. The formation of the object is achieved through the employment of a hydroelastomer press of the QFC (Quintus Technologies) type. The forming process is simulated using PAM-STAMP software, developed by the French company ESI Group. The objective of the present study is to ascertain the optimal configuration of the blank by optimizing the discrepancy between the dynamic calculations and the design model using sweep contours. The resulting new shape of the part allows for the formation of parts with minimal deviation from their design contours. Full article

This study proposes the use of a low-symmetry hexagonal photonic crystal (LSHPC) incorporating Kerr-type nonlinearity to enhance self-collimation. The equifrequency contours (EFCs) of a C₂-symmetric LSHPC composed of nonlinear LiNbO₃ rods are analyzed as a function of the nonlinear refractive index. The self-collimation characteristics, transmission spectrum, group velocity dispersion (GVD), and third-order dispersion (TOD) are investigated using the Plane Wave Expansion (PWE) and Finite Difference Time Domain (FDTD) methods. The results demonstrate that increasing the nonlinear index leads to a significant flattening of the EFCs, which enhances self-collimation performance. Furthermore, symmetry-lowering perturbations improve beam confinement and enable all-angle self-collimation. These findings highlight the potential of Kerr-type nonlinear photonic crystals for integrated photonic circuits requiring precise control over light propagation. Full article

This study proposes an improved YOLO11n-seg instance segmentation model to address the limitations of existing models in accurately identifying mature blueberries in complex greenhouse environments. Current methods often lack sufficient accuracy when dealing with complex scenarios, such as fruit occlusion, lighting variations, and target overlap. To overcome these challenges, we developed a novel approach that integrates a Spatial–Channel Adaptive (SCA) attention mechanism and a Dual Attention Balancing (DAB) module. The SCA mechanism dynamically adjusts the receptive field through deformable convolutions and fuses multi-scale color features. This enhances the model’s ability to recognize occluded targets and improves its adaptability to variations in lighting. The DAB module combines channel–spatial attention and structural reparameterization techniques. This optimizes the YOLO11n structure and effectively suppresses background interference. Consequently, the model’s accuracy in recognizing fruit contours improves. Additionally, we introduce Normalized Wasserstein Distance (NWD) to replace the traditional intersection over union (IoU) metric and address bias issues that arise in dense small object matching. Experimental results demonstrate that the improved model significantly improves target detection accuracy, recall rate, and mAP@0.5, achieving increases of 1.8%, 1.5%, and 0.5%, respectively, over the baseline model. On our self-built greenhouse blueberry dataset, the mask segmentation accuracy, recall rate, and mAP@0.5 increased by 0.8%, 1.2%, and 0.1%, respectively. In tests across six complex scenarios, the improved model demonstrated greater robustness than mainstream models such as YOLOv8n-seg, YOLOv8n-seg-p6, and YOLOv9c-seg, especially in scenes with dense occlusions. The improvement in mAP@0.5 and F1 scores validates the effectiveness of combining attention mechanisms and multiple metric optimizations, for instance, segmentation tasks in complex agricultural scenes. Full article

Driven by the rapid development of e-commerce and intelligent logistics, the volume of express delivery services has surged, making the efficient and accurate identification of shipping information a core requirement for automatic sorting systems. However, traditional Optical Character Recognition (OCR) technology struggles to meet the accuracy and real-time demands of complex logistics scenarios due to challenges such as image distortion, uneven illumination, and field overlap. This paper proposes a three-level collaborative recognition method based on deep learning that facilitates structured information extraction through regional normalization, dual-path parallel extraction, and a dynamic matching mechanism. First, the geometric distortion associated with contour detection and the lightweight direction classification model has been improved. Second, by integrating the enhanced YOLOv5s for key area localization with the upgraded PaddleOCR for full-text character extraction, a dual-path parallel architecture for positioning and recognition has been constructed. Finally, a dynamic space–semantic joint matching module has been designed that incorporates anti-offset IoU metrics and hierarchical semantic regularization constraints, thereby enhancing matching robustness through density-adaptive weight adjustment. Experimental results indicate that the accuracy of this method on a self-constructed dataset is 89.5%, with an F1 score of 90.1%, representing a 24.2% improvement over traditional OCR methods. The dynamic matching mechanism elevates the average accuracy of YOLOv5s from 78.5% to 89.7%, surpassing the Faster R-CNN benchmark model while maintaining a real-time processing efficiency of 76 FPS. This study offers a lightweight and highly robust solution for the efficient extraction of order information in complex logistics scenarios, significantly advancing the intelligent upgrading of sorting systems. Full article