Three-dimensional Data Research Articles

Laser Scanning for Canopy Characterization in Hazelnut Trees: A Preliminary Approach to Define Growth Habitus Descriptor

The accurate definition of tree growth descriptors is a crucial step in enhancing orchard management, allowing cultivar identification within an orchard and in new genotype selection for breeding programs. In apple, almond, and olive orchards, Terrestrial Laser Scanning (TLS) technologies have been already used to identify different architectural groups, but not in hazelnut yet. This study utilized TLS to investigate the canopy structure of hazelnut trees of four different Italian varieties, with and without leaves. TLS proved to be a sensor capable of collecting three-dimensional data from hazelnut field trials and allowed the definition and selection of hazelnut plant descriptors by morphological traits and morphological indexes. Nineteen descriptors, eight morphologic traits and 11 morphological indexes have been identified as reliable suitable descriptors of hazelnut cultivar and in breeding evaluations, according to Biodiversity, FAO and CIHEAM. Many of the selected descriptors are related to the tree habit, vigour and branching density. Two useful indexes have also been defined: Canopy Uprightness (CU) Index and the Index of Canopy Opening (ICO). The descriptors allowed us to distinguish the four studied hazelnut cultivars based on their growth habit; in particular the cultivar Tonda Gentile delle Langhe showed a growth habit that is a lot different from that of the other ones.

Multiphase deformation of a multilayered salt giant: Salt tectonics in the Levant Basin

The Mediterranean salt giant deposited during the Messinian salinity crisis (5.97−5.33 Ma) is the youngest known example of its kind on Earth. Unlike older layered evaporite sequences, which typically become gravitationally unstable and pierce through overlying formations, the Mediterranean salt remains relatively undeformed at shallow depths, with its internal stratification largely preserved due to minimal postdepositional deformation. This provides an opportunity to study the early stages of salt deformation. Still, the restoration of a repeatedly deformed multilayered sequence remains a challenge, even when original stratigraphic markers are identified. To reconstruct the history of deformation, one needs to distinguish between the various deformation phases and restore the evolving structures in sequence. In practice, such a restoration is rarely possible because kinematic markers (vectors) are nearly impossible to observe after multiple overprinting deformation phases. The strength of this study is rooted in the comprehensive regional mapping of a diverse array of faults and fold crests across five stratigraphic levels. This detailed mapping revealed several dominant deformation groups, shedding light on the tectonic processes at play. By assuming that the gliding direction was approximately perpendicular to the azimuth of the faults and the fold crests, and that asymmetric verging folds hint at the direction of motion, we inferred the kinematic patterns. Additionally, our analysis of crosscutting relationships, truncation features, and variations in syntectonic thickness allowed us to reconstruct a chronological sequence of deformation events. This study identified and chronologically constrained multiple deformation phases of the Messinian salt in the Levant Basin, providing new insights into both syn- and postdepositional kinematics. Our basinwide analysis underscores the key roles of regional tilting, buttressing structures, and differential sedimentary loading in controlling salt flow. Based on an extensive three-dimensional seismic reflection data set, this work presents the first high-resolution salt tectonic analysis of its scale, advancing understanding of early-stage halokinesis and setting a benchmark for future global research in sedimentary basins.

High-speed sintering on monolithic zirconia: Effects on surface and antagonist wear in vitro.

High-speed sintering on monolithic zirconia: Effects on surface and antagonist wear in vitro.

Overlapping point cloud registration algorithm based on KNN and the channel attention mechanism

With the advancement of sensor technologies such as LiDAR and depth cameras, the significance of three-dimensional point cloud data in autonomous driving and environment sensing continues to increase.Point cloud registration stands as a fundamental task in constructing high-precision environmental models, with particular significance in overlapping regions where the accuracy of feature extraction and matching directly impacts registration quality. Despite advancements in deep learning approaches, existing methods continue to demonstrate limitations in extracting comprehensive features within these overlapping areas. This study introduces an innovative point cloud registration framework that synergistically combines the K-nearest neighbor (KNN) algorithm with a channel attention mechanism (CAM) to significantly enhance feature extraction and matching capabilities in overlapping regions. Additionally, by designing an effectiveness scoring network, the proposed method improves registration accuracy and enhances system robustness in complex scenarios. Comprehensive evaluations on the ModelNet40 dataset reveal that our approach achieves markedly superior performance metrics, demonstrating significantly lower root mean square error (RMSE) and mean absolute error (MAE) compared to established methods including iterative closest point (ICP), Robust & Efficient Point Cloud Registration using PointNet (PointNetLK), Go-ICP, fast global registration (FGR), deep closest point (DCP), self-supervised learning for a partial-to-partial registration (PRNet), and Iterative Distance-Aware Similarity Matrix Convolution (IDAM). This performance advantage is consistently maintained across various challenging conditions, including unseen shapes, novel categories, and noisy environments. Furthermore, additional experiments on the Stanford dataset validate the applicability and robustness of the proposed method for high-precision 3D shape registration tasks.

Referenceless 4Dflow cardiovascular magnetic resonance with deep learning.

Referenceless 4Dflow cardiovascular magnetic resonance with deep learning.

Functional task kinematics in older adults: The role of fear of falling and fall history as measured by themovement deviation profile.

Functional task kinematics in older adults: The role of fear of falling and fall history as measured by themovement deviation profile.

Turbulent flow modeling in tur-bines via convolutional neural networks

Accurate and rapid assessment of turbulent flows in turbine working channels remains one of the key challenges in computational gas and fluid dynamics, especially under conditions of high demands on the efficiency and reliability of turbomachinery. Traditional approaches, such as Reynolds-based simulation (RANS) or large eddy simulation (LES), although providing acceptable accuracy, are associated with high computational resource and time costs. This paper proposes an alternative approach based on the use of convolutional neural networks (CNN) as a surrogate model for reproducing three-dimensional velocity and pressure fields in turbulent flows. The developed architecture is based on a modified version of U-Net and adapted for three-dimensional input data. A comparison with LES results showed that the proposed model is capable of reproducing key flow characteristics, including vortex structure and pressure gradients, with a high degree of accuracy. At the same time, a significant acceleration of calculations is achieved — up to 10³ times compared to classical numerical methods. The proposed neural network model demonstrates stability to changes in geometric parameters and can be easily reconfigured for other channel configurations. The results obtained highlight the potential of deep learning in turbulent flow modelling and open up prospects for the integration of such models into real-time engineering calculations.

Convolutional neural network for automated quantitative analysis of non-destructively acquired three-dimensional corrosion pit morphology data

Convolutional neural network for automated quantitative analysis of non-destructively acquired three-dimensional corrosion pit morphology data

Quantifying axonal features of human superficial white matter from three-dimensional multibeam serial electron microscopy data assisted by deep learning.

Quantifying axonal features of human superficial white matter from three-dimensional multibeam serial electron microscopy data assisted by deep learning.

IVR-GS: Inverse Volume Rendering for Explorable Visualization via Editable 3D Gaussian Splatting.

In volume visualization, users can interactively explore the three-dimensional data by specifying color and opacity mappings in the transfer function (TF) or adjusting lighting parameters, facilitating meaningful interpretation of the underlying structure. However, rendering large-scale volumes demands powerful GPUs and high-speed memory access for real-time performance. While existing novel view synthesis (NVS) methods offer faster rendering speeds with lower hardware requirements, the visible parts of a reconstructed scene are fixed and constrained by preset TF settings, significantly limiting user exploration. This article introduces inverse volume rendering via Gaussian splatting (iVR-GS), an innovative NVS method that reduces the rendering cost while enabling scene editing for interactive volume exploration. Specifically, we compose multiple iVR-GS models associated with basic TFs covering disjoint visible parts to make the entire volumetric scene visible. Each basic model contains a collection of 3D editable Gaussians, where each Gaussian is a 3D spatial point that supports real-time scene rendering and editing. We demonstrate the superior reconstruction quality and composability of iVR-GS against other NVS solutions (Plenoxels, CCNeRF, and base 3DGS) on various volume datasets. The code is available at https://github.com/TouKaienn/iVR-GS.

Impact of twin shield tunnel construction on the deformation of bridge pile groups beneath multiple bridges: a case study

With the rapid advancement of urbanization, the demand for efficient transportation networks, particularly for shield tunnels, has been rising. However, tunnel excavation significantly affects the surrounding soil and elevated bridge pile foundations, potentially leading to complex mechanical responses and safety risks for structures. Although prior studies have examined the impact of tunnels on pile foundations, systematic research on the deformation patterns of pile groups remains limited. Based on the case of the Zhengzhou Metro Line five crossing beneath the Shijiazhuang-Wuhan High-Speed Railway Bridge, this study uses three-dimensional finite element numerical simulations and field monitoring data to analyze the effects of tunnel excavation on bridge pile groups. The results reveal that: (1) For single-track excavation, the maximum horizontal displacement of the pile group increases to 1.587 mm, while for double-track excavation, it reaches 1.813 mm, with pile settlement increases to a maximum of 5.03 mm; (2) The maximum settlement and horizontal deformation of the bridge piles show a negative correlation with the minimum spacing between piles, with deformation increasing exponentially as the spacing decreases; (3) The excavation process at all stages affects the piers, with initial maximum settlement increasing from 0.787 mm to 3.569 mm, a growth rate of 201.91%, and maximum horizontal displacement reaching 1.576 mm. The final settlement and horizontal displacement are 5.171 mm and 1.770 mm, respectively; (4) Field monitoring and numerical analysis indicate that bridge piles in close proximity (e.g., piles 17#, 18#, 35#) are more significantly impacted by the tunnel, with maximum settlement and inclination reaching 4.5 mm and 0.008, while more distant piles (e.g., piles 16#, 20#, 34#) are less affected. These findings offer important insights for the design and construction of urban infrastructure projects, contributing to the sustainable development of cities by ensuring the safety and stability of built environment within rapidly evolving urban social spaces.

Masked Convolutions Within Skip Connections for Video Anomaly Detection

Video anomaly detection plays a crucial role in various fields such as surveillance, health monitoring, and industrial quality control. This research paper introduces a novel contribution to the field by presenting MaskedConv3D layers within a modified UNet architecture. These MaskedConv3D layers provide a unique approach to information propagation in three-dimensional video data by selectively masking temporal regions of convolutional kernels. By incorporating these layers into the skip connections of the UNet, the model gains the ability to infer missing information in the temporal domain based on the surrounding context. This innovative mechanism enhances the preservation of spatial and temporal details, addressing the challenge of effectively detecting anomalies in video data. The proposed methodology is evaluated on popular video datasets, showcasing its effectiveness in capturing intricate patterns and contexts. The results highlight the superiority of the modified UNet with MaskedConv3D layers compared to traditional approaches. Overall, this research introduces a novel technique for information propagation in video data and demonstrates its potential for advancing video anomaly detection.

Visuohaptic Feedback in Robotic-Assisted Spine Surgery for Pedicle Screw Placement.

Introduction: Robotic-assisted (RA) spine surgery enhances pedicle screw placement accuracy through real-time navigation and trajectory guidance. However, the absence of traditional direct haptic feedback by freehand instrumentation remains a concern for some, particularly in minimally invasive (MIS) procedures where direct visual confirmation is limited. During RA spine surgery, navigation systems display three-dimensional data, but factors such as registration errors, intraoperative motion, and anatomical variability may compromise accuracy. This technical note describes a visuohaptic intraoperative phenomenon observed during RA spine surgery, its underlying mechanical principles, and its utility. During pedicle screw insertion with a slow-speed automated drill in RA spine procedures, a subtle and rhythmic variation in resistance has been observed both visually on the navigation interface and haptically through the handheld drill. This intraoperative pattern is referred to in this report as a cyclical insertional torque (CIT) pattern and has been noted across multiple cases. The CIT pattern is hypothesized to result from localized stick-slip dynamics, where alternating phases of resistance and release at the bone-screw interface generate periodic torque fluctuations. The pattern is most pronounced at low insertion speeds and diminishes with increasing drill velocity. CIT is a newly described intraoperative observation that may provide visuohaptic feedback during pedicle screw insertion in RA spine surgery. Through slow-speed automated drilling, CIT offers a cue for bone engagement, which could support intraoperative awareness in scenarios where tactile feedback is reduced or visual confirmation is indirect. While CIT may enhance surgeon confidence during screw advancement, its clinical relevance, reproducibility, and impact on placement accuracy have yet to be validated.

Estimating tree biomass using UAV Lidar in Braila Islands LTSER site

Forest biomass provides a valuable metric for assessing both ecosystem productivity and carbon storage capabilities. Unmanned aerial vehicle (UAV) equipped with Light Detection and Ranging (LiDAR) technology, represents a powerful tool for estimating tree biomass due to its capacity to produce high-resolution three-dimensional data. We tested a methodology for estimating tree biomass within an area located in the flood plain of the Danube River, part of Brăila Islands Long-Term Socio-Ecological Research (LTSER) site. This area is characterized by floodplain forest plantations and natural wetlands under semi-natural hydrological regime. We selected six forest plots, representative for the area in terms of species composition and age structure. The forest composition is mainly based on willow (Salix sp.), poplar (Populus sp.) and ash (Fraxinus sp) species, planted in relation with the flooding regime. On the ground, diameter at breast height (DBH), and total tree height for 80 individual trees were recorded in order to estimate their biomass using species-specific allometric equations. These estimations are further used for LIDAR based model calibration. We have conducted a LiDAR survey over the sampling area of approx. 22 Ha, collecting a high-density point cloud to capture forest structure. The LiDAR dataset is used to generate a Digital Terrain Model (DTM) and a Canopy Height Model (CHM), as well as to derive forest and tree specific indicators (i.e. density, number of stems, heights). Preliminary results indicate that canopy height and LiDAR-derived metrics exhibit good correlations with field-based biomass estimates. Additionally, integrating LiDAR-derived, forest structure with environmental covariates, such as soil moisture and topographic features, improves biomass estimation accuracy. The study highlights the potential of UAV LiDAR, for biomass assessment in floodplain forests, in order to improve traditional field-based inventory methods that are challengingdue to seasonal flooding and dense vegetation and accessibility. The findings have implications for carbon stock assessments, ecosystem monitoring, and sustainable management strategies within the Brăila Islands LTSER site.

Production Process Improvement Towards Lean Process Effectiveness using FlexSim Simulation Software

Lean involves optimizing entire value streams to minimize waste and enhance customer value in production practices. Process improvement encompasses the identification, analysis and enhancement of existing business processes to optimize performance and align with best practice standards for a company's products or services. Companies face the imperative of increasing production to meet demand per day. The challenges include enhancing the process effectiveness in the production line. Process effectiveness measures a process's ability to produce desired results, evaluated qualitatively. FlexSim integrates simulation, artificial intelligence, three-dimensional computer image processing and data processing technology. The research objectives include identifying the operational process flow in the study company, proposing a new production process flow using FlexSim simulation software and evaluating the process flow for reducing lead time as part of process improvement. This research adopts a qualitative method, collecting data through observation and interviews. This research helps companies to increase the efficiency and effectiveness of their productivity. The data is analysed using simulation software, focusing on production process improvement to enhance process effectiveness in the PMI's company. The study's outcome reveals a 19.92% reduction in lead time achieved by consolidating processes and eliminating waste, such as waiting time, transportation and product motion. This reduction translates to cost savings and increased profitability, highlighting the positive impact of the implemented changes on the overall performance of the company.

Mixed image detection method of belt coal blockage and leakage based on improved RetinaNet mode

The mixed image characteristics of coal blocking and coal leakage by belt conveying are complex, and the current identification method lacks timeliness. A hybrid image detection method based on improved RetinaNet model is proposed. This method uses CCD camera and lidar to construct a hybrid imaging system to realize three-dimensional imaging of coal conveying belt. An improved RetinaNet model is used to train the labeled three-dimensional image data set, and features are extracted and fused through attention mechanism and feature pyramid to detect coal blockage and coal leakage, and it is deployed in production environment to realize real-time monitoring. The test results show that this method can comprehensively collect the images of coal conveying by belt, and accurately identify the detection of coal blockage and coal leakage. This method is applied to the actual production environment, which can monitor the situation of belt coal transportation in real time and accurately detect coal blockage and leakage, which is of great significance to improve the production efficiency of coal transportation, reduce economic losses and ensure production safety.

Multiscale hierarchy denoising method for heritage building point cloud model noise removal

In recent years, interest in the three-dimensional (3D) data documentation of heritage buildings has been growing. The collection of detailed and accurate 3D point-cloud information by acquiring heritage-building data has facilitated various applications. These applications encompass historical architectural information retrieval, preservation and monitoring, augmented reality, virtual reality, and the generation of heritage building information models. Point clouds originate from 3D scanners, human–computer interactions, and other devices exposed to unnecessary environments. In this context, point clouds are inevitably affected by noise and outliers. Factors contributing to noise include the limitations of sensors, device defects, and the illumination or reflection characteristics of the studied objects. Thus, addressing noise and outliers presents a challenge when storing point cloud data. Denoising is a critical step in data processing for point clouds when applied to heritage architecture. The accuracy of the point cloud model in heritage architecture is highly dependent on noise and outliers. This study proposes a multiscale hierarchy denoising method, the process of which is as follows. First, we divided the point cloud model of heritage architecture according to the architectural structure. The density-based spatial clustering of applications with noise algorithm was then used to perform large-scale point cloud denoising. For small-scale noise, denoising is achieved on a macroscopic basis by systematically removing noise and outliers from the heritage architectural point cloud model using statistical and bilateral filtering techniques. This process improves the quality and accuracy of point cloud data related to heritage buildings.

Limited distribution of cratonic strike-slip faults in Sichuan basin and their genesis mechanisms

IntroductionThe genesis mechanism of cratonic strike-slip faults in the Sichuan basin is not yet fully resolved, largely due to the lack of comprehensive characterization of fault distribution, which constrains the understanding of the relationship between fault development and tectonic events.MethodIn this study, we supplement extensive three-dimensional seismic data from areas beyond central Sichuan to delineate fault distribution across the entire basin. By integrating these findings with regional tectonic processes, the genesis mechanisms of these faults are proposed.ResultThe results reveal that cratonic strike-slip faults are exclusively concentrated in central Sichuan and the Weiyuan area, with no such faults observed elsewhere in the basin. This spatial restriction suggests that their formation was governed by the unique tectonic evolution of central Sichuan.DiscussionFurthermore, three lines of evidence prove that multi-stage uplifts played a pivotal role in fault development: (1) Faults of varying orientations exhibit distinct development periods, aligning with regional uplift phases; (2) Fault orientations correlate with the bending directions of strata induced by uplifts; and (3) Additional examples of analogous faults in Jurassic strata near the foreland front uplift further corroborate the link between fault orientation and strata bending caused by uplift processes. The central Sichuan region, characterized by superimposed multi-phase paleo-uplift evolution, thus serves as the focal zone for these faults. It is also inferred that the so-called “cratonic strike-slip” fault system in central Sichuan is not a genuine strike-slip system but rather a composite of normal faults with different directions and periods. The genesis of these normal faults is attributed to the tensile strain resulting from strata bending during uplift events. This study not only provides a more comprehensive characterization of fault distribution but also introduces a novel genetic model for these faults, which provides crucial data and new ideas for improving the theory of the fault system within the basins.

Transforming 3D MRI to 2D Feature Maps Using Pre-Trained Models for Diagnosis of Attention Deficit Hyperactivity Disorder.

Background: According to the World Health Organization (WHO), approximately 5% of children and 2.5% of adults suffer from attention deficit hyperactivity disorder (ADHD). This disorder can have significant negative consequences on people's lives, particularly children. In recent years, methods based on artificial intelligence and neuroimaging techniques, such as MRI, have made significant progress, paving the way for development of more reliable diagnostic tools. In this proof of concept study, our aim was to investigate the potential utility of neuroimaging data and clinical information in combination with a deep learning-based analytical approach, more precisely, a novel feature extraction technique for the diagnosis of ADHD with high accuracy. Methods: Leveraging the ADHD200 dataset, which encompasses demographic information and anatomical MRI scans collected from a diverse ADHD population, our study focused on developing modern deep learning-based diagnostic models. The data preprocessing employed a pre-trained Visual Geometry Group16 (VGG16) network to extract two-dimensional (2D) feature maps from three-dimensional (3D) anatomical MRI data to reduce computational complexity and enhance diagnostic power. The inclusion of personal attributes, such as age, gender, intelligence quotient, and handedness, strengthens the diagnostic models. Four deep-learning architectures-convolutional neural network 2D (CNN2D), CNN1D, long short-term memory (LSTM), and gated recurrent units (GRU)-were employed for analysis of the MRI data, with and without the inclusion of clinical characteristics. Results: A 10-fold cross-validation test revealed that the LSTM model, which incorporated both MRI data and personal attributes, had the best diagnostic performance among all tested models in the diagnosis of ADHD with an accuracy of 0.86 and area under the receiver operating characteristic (ROC) curve (AUC) score of 0.90. Conclusions: Our findings demonstrate that the proposed approach of extracting 2D features from 3D MRI images and integrating these features with clinical characteristics may be useful in the diagnosis of ADHD with high accuracy.

Parallel B-Scan YOLO Plus A-Scan Voting Method for Pavement Integrity Assessment Based on Three-Dimensional Ground-Penetrating Radar Data

Pavement integrity assessment based on ground-penetrating radar (GPR) remains challenging because of the uncertainty of the relationship between pavement damage and GPR reflection patterns. Further, existing image-processing techniques for analyzing GPR data still need significant improvements. In this study, a novel framework termed “parallel B-scan You Only Look Once (YOLO) plus A-scan voting” (P-BYPAV) was proposed to improve the performance of YOLO classification in the presence of ambiguous B-scan image patterns. The framework was developed by utilizing GPR-aided coring and falling-weight deflectometer test data. GPR B-scan reflection patterns can be classified into four categories: (1) the valley-peak-valley pattern (V-P-V) representing intact and healthy interfaces where moisture damage is minimal; (2) the peak-valley-peak pattern (P-V-P) representing the presence of air voids at layer interfaces, with a strong correlation to severe debonding; (3) the amplified valley-peak-valley pattern (Amp. V-P-V) representing the presence of moisture at layer interfaces, corresponding to various level of pavement damages; and (4) the symmetric peak-valley pattern, which indicates ambiguity in distinguishing moisture content and air voids. A dataset of 300 labeled images was created using the knowledge of these patterns. The dataset was then used to train the proposed P-BYPAV framework, demonstrating that P-BYPAV successfully integrated the advantages of YOLO and A-scan data recognition. The results show that P-BYPAV resolved more than 97% of the ambiguous regions outlined by YOLO alone.