Search Results (482)

The reliability of electric vehicle (EV) batteries requires detecting surface defects but also precisely locating them on the physical module for automated inspection, repair, and process optimisation. Conventional 2D computer vision methods, though accurate in image-space, do not provide traceable, real-world defect coordinates on complex or curved battery surfaces, limiting utility for digital twins, root cause analysis, and automated quality control. This work proposes a hybrid inspection pipeline that produces millimetre-level three-dimensional (3D) defect maps for EV battery modules. The approach integrates (i) calibrated dual-view multi-view geometry to project defect points onto the CAD geometry and triangulate them where dual-view coverage is available, (ii) single-image neural 3D shape inference calibrated to the module geometry to complement regions with limited multi-view coverage, and (iii) generative, physically informed augmentation of rare or complex defect types. Defects are first detected in 2D images using a convolutional neural network (CNN), then projected onto a dense 3D CAD model of each module, complemented by a single-image depth prediction in regions with limited dual-view coverage, yielding true as-built localisation on the battery’s surface. GenAI methods are employed to expand the dataset with synthetic defect variations. Synthetic, physically informed defect examples are incorporated during training to mitigate the scarcity of rare defect types. Evaluation on a pilot industrial dataset, with a physically measured reference subset, demonstrates that the hybrid 3D approach achieves millimetre-scale localisation accuracy and outperforms a per-view CNN baseline in both segmentation and 3D continuity. Full article

Inspecting the inner walls of large pressure vessels requires accurate weld seam recognition, complete coverage, and precise path tracking, particularly in low-feature environments. This paper presents a fully autonomous mobile robotic system that integrates weld seam detection, localization, and tracking to support ultrasonic testing. An improved Differentiable Binarization Network (DBNet) combined with the Spatially Variant Transformer (SVTR) model enhances digital stamp recognition, while weld paths are reconstructed from three-dimensional position data acquired via binocular stereo vision. To ensure complete traversal and accurate tracking, a global–local hierarchical planning strategy is implemented: the A-star (A*) algorithm performs global path planning, the Rapidly Exploring Random Tree Connect (RRT-Connect) algorithm handles local path generation, and point cloud normal–based spherical interpolation produces smooth tracking trajectories for robotic arm motion control. Experimental validation demonstrates a 94.7% digital stamp recognition rate, 95.8% localization success, 1.65 mm average weld tracking error, 2.12° normal fitting error, 98.2% seam coverage, and a tracking speed of 96 mm/s. These results confirm the system’s capability to automate weld seam inspection and provide a reliable foundation for subsequent ultrasonic testing in pressure vessel applications. Full article

Automated three-dimensional plant phenotyping is an essential tool for non-destructive analysis of plant growth and structure. This paper presents a low-cost system based on stereo vision for depth estimation and morphological characterization of maize plants. The system incorporates an automatic detection stage for the object of interest using deep learning techniques to delimit the region of interest (ROI) corresponding to the plant. The Semi-Global Block Matching (SGBM) algorithm is applied to the detected region to compute the disparity map and generate a partial three-dimensional representation of the plant structure. The ROI delimitation restricts the disparity calculation to the plant area, reducing processing of the background and optimizing computational resource use. The deep learning-based detection stage maintains stable foliage identification even under varying lighting conditions and shadowing, ensuring consistent depth data across different experimental conditions. Overall, the proposed system integrates detection and disparity estimation into an efficient processing flow, providing an accessible alternative for automated three-dimensional phenotyping in agricultural environments. Full article

Melanoma is among the most lethal forms of skin cancer, where early and accurate diagnosis significantly improves patient survival. Traditional diagnostic pathways, including clinical inspection and dermoscopy, are constrained by interobserver variability and limited access to expertise. Between 2020 and 2025, advances in artificial intelligence (AI) and medical imaging technologies have substantially redefined melanoma diagnostics. This narrative review synthesizes key developments in AI-based approaches, emphasizing the progression from convolutional neural networks to vision transformers and multimodal architectures that incorporate both clinical and imaging data. We examine the integration of AI with non-invasive imaging techniques such as reflectance confocal microscopy, high-frequency ultrasound, optical coherence tomography, and three-dimensional total body photography. The role of AI in teledermatology and mobile applications is also addressed, with a focus on expanding diagnostic accessibility. Persistent challenges include data bias, limited generalizability across diverse skin types, and a lack of prospective clinical validation. Recent regulatory frameworks, including the European Union Artificial Intelligence Act and the United States Food and Drug Administration’s guidance on adaptive systems, are discussed in the context of clinical deployment. The review concludes with perspectives on explainable AI, federated learning, and strategies for equitable implementation in dermatological oncology. Full article

For large curved parts with complex surfaces, which often exhibit both symmetry and asymmetry in their geometric features, the multi-view combined with the phase shift method and highlight regions treatment method has been proposed and applied to the online measurement system. The hardware components of the measuring system include a self-designed multi-vision platform and a multi-view three-dimensional measurement platform composed of rotating platform, robot and linear guide rail. The overall calibration of the system was conducted to guarantee the effectiveness of the measurement point cloud splicing of each viewing angle. And the system integrates the three-dimensional measurement technology of multi vision combined with the phase shift method and online measure system to realize full coverage and high-precision measurement of the impeller—addressing both its inherent symmetry (regular blade arrangement) and local asymmetry (irregular edge details)—and controls the relative error of the measured size and the actual size within 1%. In addition, the highlight regions treatment method has also been proposed. By adjusting the camera’s exposure time to change the light intensity of the captured images, images under different exposures and their valid pixels are obtained, thereby facilitating the synthesis of a composite image free of highlight phenomena. Experimental results demonstrate that the proposed method can achieve full-coverage measurement of the measured object and effective measurement of highlight regions. Full article

Three-dimensional (3D) reconstruction is increasingly being adopted in construction site management. While most existing studies rely on auxiliary equipment such as LiDAR and depth cameras, monocular depth estimation offers broader applicability under typical site conditions, yet it has received limited attention due to the inherent scale ambiguity in monocular vision. To address this limitation, this study proposes a safety helmet-based scale recovery framework that enables low-cost, monocular 3D reconstruction for construction site monitoring. The method utilizes safety helmets as exemplary scale carriers due to their standardized dimensions and frequent appearance in construction scenes. A Standard Template Library (STL) comprising multi-angle safety helmet masks and dimensional information is established and linked to site imagery through template matching. Following three-dimensional scale recovery, multi-frame fusion is applied to optimize the scale factors. Experimental results on multiple real construction videos demonstrate that the proposed method achieves high reconstruction accuracy, with a mean relative error below 10% and outliers not exceeding 5%, across diverse construction environments without site-specific calibration. This work aims to contribute to the practical application of monocular vision in engineering management by leveraging ubiquitous site objects as metrological references. Full article

For human–robot gesture-driven control, this paper proposes a deep learning-based approach that employs both static and dynamic gestures to drive and control robots for object-grasping and delivery tasks. The method utilizes two-dimensional Convolutional Neural Networks (2D-CNNs) for static gesture recognition and a hybrid architecture combining three-dimensional Convolutional Neural Networks (3D-CNNs) and Long Short-Term Memory networks (3D-CNN+LSTM) for dynamic gesture recognition. Results on a custom gesture dataset demonstrate validation accuracies of 95.38% for static gestures and 93.18% for dynamic gestures, respectively. Then, in order to control and drive the robot to perform corresponding tasks, hand pose estimation was performed. The MediaPipe machine learning framework was first employed to extract hand feature points. These 2D feature points were then converted into 3D coordinates using a depth camera-based pose estimation method, followed by coordinate system transformation to obtain hand poses relative to the robot’s base coordinate system. Finally, an experimental platform for human–robot gesture-driven interaction was established, deploying both gesture recognition models. Four participants were invited to perform 100 trials each of gesture-driven object-grasping and delivery tasks under three lighting conditions: natural light, low light, and strong light. Experimental results show that the average success rates for completing tasks via static and dynamic gestures are no less than 96.88% and 94.63%, respectively, with task completion times consistently within 20 s. These findings demonstrate that the proposed approach enables robust vision-based robotic control through natural hand gestures, showing great prospects for human–robot collaboration applications. Full article

Background/Objectives: Children with visual impairments (VI) experience lower Quality of Life (QoL), higher sedentary time, and reduced motor competence as compared to their sighted peers, posing challenges to their health, well-being, and psychosocial development. While several QoL instruments have been developed internationally for children/youth with VI, none have been validated for use with U.S. pediatric populations. The purpose of this study was to evaluate the psychometric properties of the VISIONS QL assessment tool tailored for children/youth with VI, with a primary aim of variable/item reduction to develop a streamlined version of the instrument. Methods: This study featured a cross-sectional, descriptive analytic design with convenience sampling. Participants were children and youth with VI, aged 9–19 years, (N = 148; Boys = 71, Girls = 77; M_age = 14.49, SD = 3.36 years). A principal components analysis (PCA) with orthogonal varimax rotation was conducted to reduce dimensionality and identify components. Results: Results of the PCA yielded three components explaining 46% of the variance: Educational Opportunities = 7 items; Social and Familial Implications = 8 items; Communication = 5 items. Overall, the VISIONS QL Brief had a high level of internal consistency reliability (α = 0.857; Ω = 0.858) and test–retest reliability (ICC = 0.89, 95% CI = 0.84–0.92). The original 63-item version showed concurrent validity with the 20-item brief scale (r = 0.92, p < 0.0001). Conclusions: Findings affirm the multidimensional nature of QoL and support the usage of the VISIONS QL Brief and its utility in settings where time, accessibility, and cognitive load are critical considerations. Full article

As the global deployment of fifth-generation (5G) networks matures, the research community is conceptualising sixth-generation (6G) systems, projected for deployment around 2030. This article presents a comprehensive, evidence-based examination of the technological innovations and applications that characterise this transition, informed by a scoping review of 57 sources published between January 2020 and August 2025. The transition to 6G signifies a fundamental transformation from a mere communication utility to an intelligent, sensing, and globally integrated cyber-physical continuum, propelled by a strategic reassessment of the network’s societal function and the practical insights gained from the 5G era. We critically analyse the foundational physical layer technologies that facilitate this vision, including Reconfigurable Intelligent Surfaces (RIS), Terahertz (THz) communications, and the transition to Extremely Large-Scale MIMO (XL-MIMO), emphasising their interdependencies and the fundamental shift towards near-field physics. The analysis encompasses the architectural transformation necessary to address this new complexity, elucidating the principles of the AI-native network, the seamless integration of Non-Terrestrial Networks (NTN) into a cohesive three-dimensional framework, and the functional convergence of communication and sensing (ISAC). We also look at how these changes affect the real world by looking at data from trials and case studies in smart cities, intelligent transportation, and digital health. The article synthesises the overarching challenges in security, sustainability, and scalability, arguing that the path to 6G is defined by two intertwined grand challenges: building a trustworthy and sustainable network. By outlining the critical research imperatives that stem from these challenges, this work offers a holistic framework for understanding how these interconnected developments are evolving wireless networks into the intelligent fabric of a digitised and sustainable society. Full article

Repeatability accuracy is a key performance metric for robotic arms. To address limitations in existing monocular vision-based measurement methods, this study proposes a likelihood-based pose estimation approach. Our method first obtains initial pose estimates through optimized likelihood estimation, then iteratively refines depth information. By modeling the statistical characteristics of multiple observed poses, we derive a global theoretical pose. Within this framework, two-dimensional feature points are backprojected into three-dimensional space to form an observed point cloud. The error between this observed cloud and the theoretical feature point cloud is computed using the Iterative Closest Point (ICP) algorithm, enabling accurate quantification of repeatability accuracy. Based on 30 repeated trials at each of five target poses, the proposed method achieved repeatability positioning accuracy of 0.0115 mm, 0.0121 mm, 0.0068 mm, 0.0162 mm, and 0.0175 mm at the five poses, respectively, with a mean value of 0.0128 mm and a standard deviation of 0.0038 mm across the poses. Compared with two existing monocular vision-based methods, it demonstrates superior accuracy and stability, achieving average accuracy improvements of 0.79 mm and 1.06 mm, respectively, and reducing the standard deviation by over 85%. Full article

Three-dimensional Human Reconstruction from Monocular Vision is a key technology in Virtual Reality and digital humans. It aims to recover the 3D structure and pose of the human body from 2D images or video. Current methods for dynamic 3D reconstruction of the human body, which are based on monocular views, have low accuracy and remain a challenging problem. This paper proposes a fast reconstruction method based on Instant Human Model (IHM) generation, which achieves highly realistic 3D reconstruction of the human body in arbitrary poses. First, the efficient dynamic human body reconstruction method, InstantAvatar, is utilized to learn the shape and appearance of the human body in different poses. However, due to its direct use of low-resolution voxels as canonical spatial human representations, it is not possible to achieve satisfactory reconstruction results on a wide range of datasets. Next, a voxel occupancy grid is initialized in the A-pose, and a voxel attention mechanism module is constructed to enhance the reconstruction effect. Finally, the Instant Human Model (IHM) method is employed to define continuous fields on the surface, enabling highly realistic dynamic 3D human reconstruction. Experimental results show that, compared to the representative InstantAvatar method, IHM achieves a 0.1% improvement in SSIM and a 2% improvement in PSNR on the PeopleSnapshot benchmark dataset, demonstrating improvements in both reconstruction quality and detail. Specifically, IHM, through voxel attention mechanisms and Mesh adaptive iterative optimization, achieves highly realistic 3D mesh models of human bodies in various poses while ensuring efficiency. Full article

This paper presents a federated learning (FL) architecture tailored for anomaly detection in wire arc additive manufacturing (WAAM) that preserves data privacy while enabling secure and distributed model training across heterogeneous process units. WAAM’s inherent process complexity, characterized by high-dimensional and asynchronous sensor streams, including current, voltage, travel speed, and visual bead profiles, necessitates a decentralized learning paradigm capable of handling non-identical client distributions without raw data pooling. To this end, the proposed framework integrates reversible data hiding in the encrypted domain (RDHE) for the secure embedding of sensor-derived features into weld images, enabling confidential parameter transmission and tamper-evident federation. Each client node employs a domain-specific long short-term memory (LSTM)-based classifier trained on locally curated time-series or vision-derived features, with model updates embedded and transmitted securely to a central aggregator. Three FL strategies, FedAvg, FedProx, and FedPer, are systematically evaluated against four robust aggregation techniques, including KRUM, Multi KRUM, and Trimmed Mean, across 100 communication rounds using eight non-independent and identically distributed (non-IID) WAAM clients. Experimental results reveal that FedPer coupled with Trimmed Mean delivers the optimal configuration, achieving maximum F1-score (0.912), area under the curve (AUC) (0.939), and client-wise generalization stability under both geometric and temporal noise. The proposed approach demonstrates near-lossless RDHE encoding (PSNR > 90 dB) and robust convergence across adversarial conditions. By embedding encrypted intelligence within weld imagery and tailoring FL to WAAM-specific signal variability, this study introduces a scalable, secure, and generalizable framework for process monitoring. These findings establish a baseline for federated anomaly detection in metal additive manufacturing, with implications for deploying privacy-preserving intelligence across smart manufacturing (SM) networks. The federated pipeline is backbone-agnostic. We instantiate LSTM clients because the sequences are short (five steps) and edge compute is limited in WAAM. The same pipeline can host Transformer/TCN encoders for longer horizons without changing the FL or security flow. Full article

Thailand’s mango export industry faces significant challenges in meeting stringent international quality standards, particularly the costly phytosanitary X-ray irradiation process. Current fixed-dose irradiation methods result in substantial energy waste due to variations in fruit size. This research presents a low-cost, real-time system that integrates computer vision and artificial intelligence (AI) to optimize this process. By capturing a single top-view 2D image, the system accurately estimates the three-dimensional characteristics (width, height, and depth) of ‘Nam Dok Mai Si Thong’ mangoes. This dimensional data is crucial for dynamically adjusting the radiation dose for each fruit, leading to significant reductions in energy consumption and operational costs. Our novel approach utilizes a Linear Regression combined with Co-Kriging (LR + CoK) model to precisely estimate fruit depth from 2D data, a key limitation in previous studies. The system demonstrated high efficacy, achieving a dimensional estimation error (RMSE) of less than 0.46 cm and a size grading accuracy of up to 93.33 percent. This technology not only enhances sorting and grading efficiency but also offers a practical solution to lower the economic and energy burden of phytosanitary treatments, directly improving the sustainability of fruit export operations. Full article

While recent industrial automation trends emphasize the importance of non-destructive inspection by material-identifying millimeter-wave, terahertz-wave, and infrared (MMW, THz, IR) monitoring, fundamental tools in these wavelength bands (such as sensors) are still immature. Although inorganic semiconductors serve as diverse sensors with well-established large-scale fine-processing fabrication, the use of those devices is insufficient for non-destructive monitoring due to the lack of photo-absorbent properties for such major materials in partial regions across MMW–IR wavelengths. To satisfy the inherent advantageous non-destructive MMW–IR material identification, ultrabroadband operation is indispensable for photo-sensors under compact structure, flexible designability, and sensitive performances. This review then introduces the recent advances of carbon nanotube film-based photo-thermoelectric imagers regarding usable and high-yield device fabrication techniques and scientific synergy among computer vision to collectively satisfy material identification with three-dimensional (3D) structure reconstruction. This review synergizes material science, printable electronics, high-yield fabrication, sensor devices, optical measurements, and imaging into guidelines as functional non-destructive inspection platforms. The motivation of this review is to introduce the recent scientific fusion of MMW–IR sensors with visible-light computer vision, and emphasize its significance (non-invasive material-identifying sub-millimeter-resolution 3D-reconstruction with 660 nm–1.15 mm-wavelength imagers at noise equivalent power within 100 pWHz^−1/2) among the existing testing methods. Full article

The safety and reliability of modern power systems are increasingly challenged by adverse environmental conditions. (1) Background: Ice accumulation on power transmission lines is recognized as a severe threat to grid stability, as tower collapse, conductor breakage, and large-scale outages may be caused, thereby making accurate monitoring essential. (2) Methods: A rule-driven and interpretable stereo vision framework is proposed for three-dimensional (3D) detection and quantitative measurement of transmission line icing. The framework consists of three stages. First, adaptive preprocessing and segmentation are applied using multiscale Retinex with nonlinear color restoration, graph-based segmentation with structural constraints, and hybrid edge detection. Second, stereo feature extraction and matching are performed through entropy-based adaptive cropping, self-adaptive keypoint thresholding with circular descriptor analysis, and multi-level geometric validation. Third, 3D reconstruction is realized by fusing segmentation and stereo correspondences through triangulation with shape-constrained refinement, reaching millimeter-level accuracy. (3) Result: An accuracy of 98.35%, sensitivity of 91.63%, specificity of 99.42%, and precision of 96.03% were achieved in contour extraction, while a precision of 90%, recall of 82%, and an F1-score of 0.8594 with real-time efficiency (0.014–0.037 s) were obtained in stereo matching. Millimeter-level accuracy (Mean Absolute Error: 1.26 mm, Root Mean Square Error: 1.53 mm, Coefficient of Determination = 0.99) was further achieved in 3D reconstruction. (4) Conclusions: Superior accuracy, efficiency, and interpretability are demonstrated compared with two existing rule-based stereo vision methods (Method A: ROI Tracking and Geometric Validation Method and Method B: Rule-Based Segmentation with Adaptive Thresholding) that perform line icing identification and 3D reconstruction, highlighting the framework’s advantages under limited data conditions. The interpretability of the framework is ensured through rule-based operations and stepwise visual outputs, allowing each processing result, from segmentation to three-dimensional reconstruction, to be directly understood and verified by operators and engineers. This transparency facilitates practical deployment and informed decision making in real world grid monitoring systems. Full article