Point Cloud Datasets Research Articles

Improved building MEP systems semantic segmentation in point clouds using a novel multi-class dataset and local–global vector transformer network

Point cloud semantic segmentation for mechanical, electrical, and plumbing (MEP) systems is crucial for establishing MEP systems digital twins. Deep learning has shown promise in automatically deriving point-wise semantic labels from point clouds. However, the utilization of 3D deep learning methods for directly segmenting complex, occluded, and densely distributed multi-class MEP systems in common building point clouds remains unexplored due to the lack of specialized public datasets. To address this, our study introduces a public, richly annotated, and multi-class point cloud dataset for MEP systems in common buildings, consisting of 92.10 million points across four buildings, 56 areas, and nine component categories. Subsequently, we propose Trans2Net, a novel and customized 3D deep learning method for the automatic and precise segmentation of MEP components, featuring local vector transformer module, global-aid vector transformer module, and learnable feature fusion operator that enhance the model’s sensitivity to local features, capture global context with sparse points, and dynamically fuse weighted local and global features. Experimental results demonstrate that Trans2Net significantly outperforms seven other state-of-the-art methods on the MEP dataset, achieving an overall accuracy of 97.39%, a mean accuracy of 95.45%, and a mean intersection over union (mIoU) of 90.76%. Notably, it surpasses the second-ranked method, PointMeta, by a margin of 9.91% in mIoU. This study provides a novel paradigm for automatically understanding MEP systems in point clouds and offers robust support for digital twin management of MEP systems, contributing to the prolonged lifespans of MEP systems.

Four-dimensional hemodynamic prediction of abdominal aortic aneurysms following endovascular aneurysm repair combining physics-informed PointNet and quadratic residual networks

Hemodynamic parameters can provide surveillance for the risk of complication of abdominal aortic aneurysms following endovascular aneurysm repair (EVAR). However, obtaining hemodynamic parameters through computational fluid dynamics (CFD) has disadvantages of complex operation and high computational costs. Recently proposed physics-informed neural networks offer novel solutions to solve these issues by leveraging fundamental physical conservation principles of fluid dynamics. Based on cardiovascular point datasets, we further propose an integration algorithm combining physics-informed PointNet and quadratic residual networks (PIPN-QN) that is capable of mapping sparse point clouds to four-dimensional hemodynamic parameters. The implemented workflow includes generating point cloud datasets through CFD simulation and dynamically reproducing the three-dimensional flow field in the spatial and temporal dimensions through deep learning. Compared with physics-informed PointNet (PIPN), the PIPN-QN reduces the mean square error of pressure and wall shear stress by around 32.1% and 33.1% and anticipates hemodynamic parameters in less than 2 s (14 400 times faster than CFD). To address the challenge of big data requirements, we quantify the universal flow field using a reduced number of supervision points, as opposed to the large number of point clouds generated from the CFD simulation. The PIPN-QN can meet the real-time hemodynamic parameters obtained from patients with abdominal aortic aneurysms following EVAR with higher accuracy, faster speed, and lower training costs.

Filtering of 3D point clouds using maximum likelihood algorithm

Recently, the 3D point cloud (PC) has become more popular as an innovative object representation. However, there is usually noise and outliers in the raw point cloud. It is essential to eliminate the noise from the point cloud and outlier data while maintaining the features and finer details intact. This paper presents a comprehensive method for filtering and classification point clouds using a maximum likelihood algorithm (ML). TOPCON GLS-2000 3D terrestrial laser scanners (TLS) have been used to collect the 3D PC data set; the scan range is up to 350 m. About 30 m apart from the study area. ScanMaster software has been used to import, view, and filter point cloud information. The position information of the points is linked with the training point cloud and the filtered point cloud to derive the nonlinear model using MATLAB software. To evaluate the quality of the denoising results, two error metrics have been used: the average angle (δ) and distance (Dmean) between the ground truth point and the resulting point. The experimental findings demonstrate that the suggested approach can effectively filter out background noise while improving feature preservation. The filtering and classifying technique is more effective and efficient compared to the selected filtering methods when applied to 3D point clouds containing a large number of points and a variety of natural characteristics.

Research on Student's T-Distribution Point Cloud Registration Algorithm Based on Local Features.

LiDAR offers a wide range of uses in autonomous driving, remote sensing, urban planning, and other areas. The laser 3D point cloud acquired by LiDAR typically encounters issues during registration, including laser speckle noise, Gaussian noise, data loss, and data disorder. This work suggests a novel Student's t-distribution point cloud registration algorithm based on the local features of point clouds to address these issues. The approach uses Student's t-distribution mixture model (SMM) to generate the probability distribution of point cloud registration, which can accurately describe the data distribution, in order to tackle the problem of the missing laser 3D point cloud data and data disorder. Owing to the disparity in the point cloud registration task, a full-rank covariance matrix is built based on the local features of the point cloud during the objective function design process. The combined penalty of point-to-point and point-to-plane distance is then added to the objective function adaptively. Simultaneously, by analyzing the imaging characteristics of LiDAR, according to the influence of the laser waveform and detector on the LiDAR imaging, the composite weight coefficient is added to improve the pertinence of the algorithm. Based on the public dataset and the laser 3D point cloud dataset acquired in the laboratory, the experimental findings demonstrate that the proposed algorithm has high practicability and dependability and outperforms the five comparison algorithms in terms of accuracy and robustness.

Binary segmentation of relief patterns on point clouds

Analysis of 3D textures, also known as relief patterns is a challenging task that requires separating repetitive surface patterns from the underlying global geometry. Existing works classify entire surfaces based on one or a few patterns by extracting ad-hoc statistical properties. Unfortunately, these methods are not suitable for objects with multiple geometric textures and perform poorly on more complex shapes. In this paper, we propose a neural network for binary segmentation to infer per-point labels based on the presence of surface relief patterns. We evaluated the proposed architecture on a high resolution point cloud dataset, surpassing the state-of-the-art, while maintaining memory and computation efficiency.

SaBi3d—A LiDAR Point Cloud Data Set of Car-to-Bicycle Overtaking Maneuvers

While cycling presents environmental benefits and promotes a healthy lifestyle, the risks associated with overtaking maneuvers by motorized vehicles represent a significant barrier for many potential cyclists. A large-scale analysis of overtaking maneuvers could inform traffic researchers and city planners how to reduce these risks by better understanding these maneuvers. Drawing from the fields of sensor-based cycling research and from LiDAR-based traffic data sets, this paper provides a step towards addressing these safety concerns by introducing the Salzburg Bicycle 3d (SaBi3d) data set, which consists of LiDAR point clouds capturing car-to-bicycle overtaking maneuvers. The data set, collected using a LiDAR-equipped bicycle, facilitates the detailed analysis of a large quantity of overtaking maneuvers without the need for manual annotation through enabling automatic labeling by a neural network. Additionally, a benchmark result for 3D object detection using a competitive neural network is provided as a baseline for future research. The SaBi3d data set is structured identically to the nuScenes data set, and therefore offers compatibility with numerous existing object detection systems. This work provides valuable resources for future researchers to better understand cycling infrastructure and mitigate risks, thus promoting cycling as a viable mode of transportation.

MP-DGCNN for the semantic segmentation of Chinese ancient building point clouds

Point cloud semantic segmentation is a key step in the scan-to-HBIM process. In order to reduce the information in the process of DGCNN, this paper proposes a Mix Pooling Dynamic Graph Convolutional Neural Network (MP-DGCNN) for the segmentation of ancient architecture point clouds. The proposed MP-DGCNN differs from DGCNN mainly in two aspects: (1) to more comprehensively characterize the local topological structure of points, the edge features are redefined, and distance and neighboring points are added to the original edge features; (2) based on a Multilayer Perceptron (MLP), an internal feature adjustment mechanism is established, and a learnable mix pooling operator is designed by fusing adaptive pooling, max pooling, average pooling, and aggregation pooling, to learn local graph features from the point cloud topology. To verify the proposed algorithm, experiments are conducted on the Qutan Temple point cloud dataset, and the results show that compared with PointNet, PointNet++, DGCNN, GACNet and LDGCNN, the MP-DGCNN segmentation network achieves the highest OA, mIOU and mAcc, reaching 90.19%,65.34% and 79.41%, respectively.

A Comparative Literature Review of Machine Learning and Image Processing Techniques Used for Scaling and Grading of Wood Logs

This literature review assesses the efficacy of image-processing techniques and machine-learning models in computer vision for wood log grading and scaling. Four searches were conducted in four scientific databases, yielding a total of 1288 results, which were narrowed down to 33 relevant studies. The studies were categorized according to their goals, including log end grading, log side grading, individual log scaling, log pile scaling, and log segmentation. The studies were compared based on the input used, choice of model, model performance, and level of autonomy. This review found a preference for images over point cloud representations for logs and an increase in camera use over laser scanners. It identified three primary model types: classical image-processing algorithms, deep learning models, and other machine learning models. However, comparing performance across studies proved challenging due to varying goals and metrics. Deep learning models showed better performance in the log pile scaling and log segmentation goal categories. Cameras were found to have become more popular over time compared to laser scanners, possibly due to stereovision cameras taking over for laser scanners for sampling point cloud datasets. Classical image-processing algorithms were consistently used, deep learning models gained prominence in 2018, and other machine learning models were used in studies published between 2010 and 2018.

An Improved Large Planar Point Cloud Registration Algorithm

The traditional Iterative Closest Point (ICP) algorithm often suffers from low computational accuracy and efficiency in certain scenarios. It is highly sensitive to the initial pose, has a poor ability to resist interference, and frequently becomes trapped in local optima. Extracting feature points accurately from partially overlapping points with weak three-dimensional features, such as smooth planes or surfaces with low curvature, is challenging using only the traditional ICP algorithm for registration. This research introduces a “First Rough then Precise” registration strategy. Initially, the target position is extracted in complex environments using an improved clustering method, which simultaneously reduces the impact of environmental factors and noise on registration accuracy. Subsequently, an improved method for calculating normal vectors is applied to the Fast Point Feature Histogram (FPFH) to extract feature points, providing data for the Sample Consistency Initial Algorithm (SAC-IA). Lastly, an improved ICP algorithm, which has strong anti-interference capabilities for partially overlapping point clouds, is utilized to merge such point clouds. In the experimental section, we validate the feasibility and precision of the proposed algorithm by comparing its registration outcomes with those of various algorithms, using both standard point cloud dataset models and actual point clouds obtained from camera captures.

FLApy: A Python package for evaluating the 3D light availability heterogeneity within forest communities

Abstract Light availability (LAv) dictates a variety of biological and ecological processes across a range of spatiotemporal scales. Quantifying the spatial pattern of LAv in three‐dimensional (3D) space can promote the understanding of microclimates that are critical to fine‐scale species distribution. However, there is still a lack of tools that are robust to evaluate spatiotemporal heterogeneity of LAv in forests. Here, we propose the Forest Light Analyzer python package (FLApy), an open‐source computational tool designed for the analysis of intra‐forest LAv variation across multiple spatial scales. FLApy is freely invoked by Python, facilitating the processing of LiDAR point cloud data into a 3D data container constructed by voxels, as well as traversal calculations related to the LAv regime by high performance synthetic hemispherical algorithm. Furthermore, FLApy incorporates 37 indicators, enabling users to expediently export and visualize LAv patterns and the evaluation of heterogeneity of LAv at two scales (voxel scale and 3D‐cluster scale) for a range of fine‐scale ecological study purposes. To validate the efficacy of the FLApy, we employed a simulated point cloud dataset that simulates forests (varying in canopy closure). Furthermore, to evaluate real world forest, we executed the standard workflow of FLApy utilizing drone‐derived data from three subtropical evergreen broad‐leaved forest dynamics plots within the Ailao Mountain Reserve. Our findings underscore that a series of indices derived from FLApy provide a robust characterization of light availability heterogeneity within diverse forest settings. Additionally, when juxtaposed with conventional monitoring techniques, the metrics offered by FLApy demonstrated better generality in our field assessments. FLApy offers ecologists a solution for rapid quantification of understory light 3D‐regimes across multiple scales, addressing the disparity between traditional manual approaches and the precision required for contemporary ecological studies. Moreover, FLApy provides robust support for the establishment and expansion of heterogeneity indices based on 3D micro‐environments, enhancing our understanding of the largely uncharted 3D structural patterns. Anticipated outcomes suggest that FLApy will enhance our knowledge concerning the intra‐forest climatic conditions into a 3D context, proving pivotal in the delineation of microhabitats and the development of detailed 3D‐scale species distribution models.

Compound 3D building modeling with structure-aware partition and primitive assembly from airborne laser scanning point clouds

ABSTRACT Urban digital twins and realistic three-dimensional (3D) scenes have further enhanced the ever-increasing demand for building models. Airborne Laser Scanning (ALS) point clouds are the predominant data source for 3D building reconstruction owing to their highly detailed spatial features. However, information on the reconstruction of compound buildings (i.e. buildings with diverse plane structures or various roof structures) is lacking. Existing methods often produce building models that suffer from topological inconsistencies and/or geometric errors. To address these issues, this study proposes an automatic building reconstruction approach based on ALS point cloud that offers both modeling flexibility and geometric consistency. Specifically, a structure-aware partitioning algorithm that partitions a complex building roof into several regular roof patches is incorporated. Subsequently, a primitive model library comprising representative primitives suitable for building reconstruction is created. Suitable primitive model selection and reconstruction optimization are employed to ensure the accuracy and compactness in the final model. To comprehensively evaluate the performance, two typical ALS point cloud datasets are tested. Experiments demonstrate the generated models can achieve a boundary consistency of 92% while ensuring Root Mean Square Error below 0.09 m. Comparative experiments with state-of-the-art methods further validate the effectiveness of the proposed approach in compound building reconstruction.

A fast point cloud registration method based on spatial relations and features

Abstract Point cloud registration plays a crucial role in mobile robot localization, map building and three-dimensional (3D) model reconstruction. However, it remains challenged by issues such as compromised accuracy and sluggish efficiency, posing significant obstacles in achieving precise and timely alignments. Therefore, we propose a lightweight and fast point cloud registration method. Firstly, we mesh the 3D point cloud, compared with the traditional gridded point cloud method, it achieves initial point cloud registration by preserving the curvature characteristics of the internal point cloud, and utilizing the spatial relationship between grid cells and the quantitative relationship between the internal point cloud. Moreover, we adopt an iterative nearest point based on KD-Tree to realize the fine registration. So, our method does not necessitate intricate feature analysis and data training, and is resilient to similar transformations, non-uniform densities and noise. Finally, we conduct point cloud registration experiments using multiple publicly available point cloud datasets and compare them with several point cloud registration methods. The results demonstrate it is able to accomplish the point cloud registration quickly and exhibit high accuracy. More importantly, it maintains its efficacy and robustness even in the presence of noisy and defective point clouds.

A Point Cloud Dataset of Vehicles Passing through a Toll Station for Use in Training Classification Algorithms

A Point Cloud Dataset of Vehicles Passing through a Toll Station for Use in Training Classification Algorithms

Development of large-scale synthetic 3D point cloud datasets for vision-based bridge structural condition assessment

Visual recognition of 3D point cloud data of bridge inspection scenes is a key step in automating the visual inspection process, which is currently largely manual and inefficient. To alleviate the lack of large-scale annotated point cloud datasets for training such 3D visual recognition algorithms, this research investigates an approach for developing large-scale synthetic point cloud datasets. The proposed approach proceeds in four steps: (1) random generation of different types of bridges in computer graphics environments; (2) sampling of camera trajectories that represent data collection scenarios during bridge inspection; (3) 3D reconstruction using Structure from Motion (SfM) applied to rendered synthetic images; (4) automated annotation of the reconstructed point cloud using ground truth masks obtained with synthetic images. Besides, this research proposes to store point uncertainty information defined by the error between the ground truth depth and the depth calculated from the SfM results. Prior to training, thresholds can be applied to this uncertainty information to control the levels of outliers in the dataset. This research demonstrates the proposed approach by generating point cloud datasets for two data collection scenarios. The effectiveness of the generated datasets is investigated by training 3D semantic segmentation algorithms and evaluating the performance on real and synthetic point cloud data. The proposed approach for point cloud dataset generation will facilitate the development of generalizable and high level-of-detail 3D recognition algorithms toward autonomous bridge inspection.

A 3D local feature description algorithm based on point distribution

It is well known that the local feature descriptor is playing an important role in 3D image recognition. But the current local feature descriptors have poor ability to describe target features in situations such as noise interference, be obscured, and changes in data resolution. On the basis of point distribution, we propose a new 3D local feature descriptor, the Point Distribution Signature Histogram(PDSH), which is relatively simple and easy to implement and has a fast identification speed by simplifying Local Reference Frame(LRF). We also rotate the local surface of the point cloud from multiple angles and project them onto three coordinate planes of the LRF. The projection planes are segmented at the same angle, and the distance information of the edge points and the count of projected points in each segment are statistically accumulated. Finally, the sub-feature vectors of the three coordinate planes are concatenated to obtain the feature descriptor of the key points. Experimental results show that our PDSH descriptor achieves a descriptiveness of 98 % on the pollution-free point cloud data set and better performance than the classical descriptors.

RailPC: A large‐scale railway point cloud semantic segmentation dataset

AbstractSemantic segmentation in the context of 3D point clouds for the railway environment holds a significant economic value, but its development is severely hindered by the lack of suitable and specific datasets. Additionally, the models trained on existing urban road point cloud datasets demonstrate poor generalisation on railway data due to a large domain gap caused by non‐overlapping special/rare categories, for example, rail track, track bed etc. To harness the potential of supervised learning methods in the domain of 3D railway semantic segmentation, we introduce RailPC, a new point cloud benchmark. RailPC provides a large‐scale dataset with rich annotations for semantic segmentation in the railway environment. Notably, RailPC contains twice the number of annotated points compared to the largest available mobile laser scanning (MLS) point cloud dataset and is the first railway‐specific 3D dataset for semantic segmentation. It covers a total of nearly 25 km railway in two different scenes (urban and mountain), with 3 billion points that are finely labelled as 16 most typical classes with respect to railway, and the data acquisition process is completed in China by MLS systems. Through extensive experimentation, we evaluate the performance of advanced scene understanding methods on the annotated dataset and present a synthetic analysis of semantic segmentation results. Based on our findings, we establish some critical challenges towards railway‐scale point cloud semantic segmentation. The dataset is available at https://github.com/NNU‐GISA/GISA‐RailPC, and we will continuously update it based on community feedback.

Pose estimation algorithm based on point pair features using PointNet + +

This study proposes an innovative deep learning algorithm for pose estimation based on point clouds, aimed at addressing the challenges of pose estimation for objects affected by the environment. Previous research on using deep learning for pose estimation has primarily been conducted using RGB-D data. This paper introduces an algorithm that utilizes point cloud data for deep learning-based pose computation. The algorithm builds upon previous work by integrating PointNet + + technology and the classical Point Pair Features algorithm, achieving accurate pose estimation for objects across different scene scales. Additionally, an adaptive parameter-density clustering method suitable for point clouds is introduced, effectively segmenting clusters in varying point cloud density environments. This resolves the complex issue of parameter determination for density clustering in different point cloud environments and enhances the robustness of clustering. Furthermore, the LineMod dataset is transformed into a point cloud dataset, and experiments are conducted on the transformed dataset to achieve promising results with our algorithm. Finally, experiments under both strong and weak lighting conditions demonstrate the algorithm's robustness.

DGPCD: a benchmark for typical official-style Dougong in ancient Chinese wooden architecture

Dougong, a distinctive component of ancient Chinese wooden architecture, holds significant importance for the preservation and restoration of such structures. In particular, the northern official-style buildings represent the pinnacle of ancient Chinese construction techniques. In the realm of cultural heritage preservation, the application of deep learning has gradually expanded, demonstrating remarkable effectiveness. Point cloud serving as a crucial source for Dougong, encapsulates various information, enabling support for tasks like Dougong point cloud classification and completion. The quality of Dougong datasets directly impacts the outcomes of DNNs (deep neural networks), as they serve as the foundational data support for these models. The typical official-style Dougong, with its standardized and repetitive structural patterns, is highly suitable for training DNNs to accurately recognize and analyze these complex architectural elements. However, due to the inherent characteristics of Dougong, such as coplanarity and occlusion, acquiring point cloud data is challenging, resulting in poor data quality and organizational difficulties. To address this issue, our study adopts a multi-source data fusion approach to tackle the challenges of insufficient data quantity and poor data quality. Further, through data augmentation, we enhance the dataset’s volume and generalize its characteristics. This effort culminates in the creation of the typical official-style Dougong Point Cloud Dataset (DG Dataset), poised to support deep learning tasks related to Dougong scenarios.

A Review of Datasets for 3D Indoor Modeling

Abstract: Indoor modeling has witnessed a substantial transformation in recent years, driven by advancements in 3D sensing technologies and data processing techniques. Point cloud datasets, representing detailed geometric information of indoor environments, play a pivotal role in enabling various applications such as architectural design, navigation, and augmented reality, and Building Information Modeling (BIM). This paper provides a comprehensive overview and comparision of the stateof-the-art datasets specifically curated for indoor modeling.

Applying machine learning to reveal the microscopic heat transfer mechanism of nanofluids as coolants

Nanofluids are considered as excellent coolants to optimize thermal management of electronic devices, where the nanoparticle morphology and the addition of surfactants can affect the thermal transport performance of nanofluids. Due to the limitations of high economic and computational cost in previous experimental and numerical simulation methods, the design of nanofluids urges for more efficient approaches. In this work, a novel machine learning framework coupled with molecular dynamics methods was proposed to model the multi-component mixing nanofluidic systems and explore the deep heat transfer mechanisms. Multi-input attribute point cloud dataset, dual channel sampling network and multi-nanoscale optimization scheme were used to improve the prediction performance of machine learning. The computational cost of the machine learning method is shortened by 36000 times compared with simulation methods. Moreover, our work can achieve up to 90 % prediction accuracy for surfactant adsorption properties. Furthermore, algorithm optimization strategy can improve the prediction accuracy of nanofluidic heat transfer performance by 40 %. The proposed framework has the potential to shorten the development cycle of nanofluidic design.