Accuracy Of Point Cloud Data Research Articles

A Triangular Grid Filter Method Based on the Slope Filter

High-precision ground point cloud data has a wide range of applications in various fields, and the separation of ground points from non-ground points is a crucial preprocessing step. Therefore, designing an efficient, accurate, and stable ground extraction algorithm is highly significant for improving the processing efficiency and analysis accuracy of point cloud data. The study area in this article was a park in Guilin, Guangxi, China. The point cloud was obtained by utilizing the UAV platform. In order to improve the stability and accuracy of the filter algorithm, this article proposed a triangular grid filter based on the Slope Filter, found violation points by the spatial position relationship within each point in the triangulation network, improved KD-Tree-Based Euclidean Clustering, and applied it to the non-ground point extraction. This method is accurate, stable, and achieves the separation of ground points from non-ground points. Firstly, the Slope Filter is used to remove some non-ground points and reduce the error of taking ground points as non-ground points. Secondly, a triangular grid based on the triangular relationship between each point is established, and the violation triangle is determined through the grid; thus, the corresponding violation points are found in the violation triangle. Thirdly, according to the three-point collinear method to extract the regular points, these points are used to extract the regular landmarks by the KD-Tree-Based Euclidean Clustering and Convex Hull Algorithm. Finally, the dispersed points and irregular landmarks are removed by the Clustering Algorithm. In order to confirm the superiority of this algorithm, this article compared the filter effects of various algorithms on the study area and filtered the 15 data samples provided by ISPRS, obtaining an average error of 3.46%. The results show that the algorithm presented in this article has high processing efficiency and accuracy, which can significantly improve the processing efficiency of point cloud data in practical applications.

Fusion of a Static and Dynamic Convolutional Neural Network for Multiview 3D Point Cloud Classification

Three-dimensional (3D) point cloud classification methods based on deep learning have good classification performance; however, they adapt poorly to diverse datasets and their classification accuracy must be improved. Therefore, FSDCNet, a neural network model based on the fusion of static and dynamic convolution, is proposed and applied for multiview 3D point cloud classification in this paper. FSDCNet devises a view selection method with fixed and random viewpoints, which effectively avoids the overfitting caused by the traditional fixed viewpoint. A local feature extraction operator of dynamic and static convolution adaptive weight fusion was designed to improve the model’s adaptability to different types of datasets. To address the problems of large parameters and high computational complexity associated with the current methods of dynamic convolution, a lightweight and adaptive dynamic convolution operator was developed. In addition, FSDCNet builds a global attention pooling, integrating the most crucial information on different view features to the greatest extent. Due to these characteristics, FSDCNet is more adaptable, can extract more fine-grained detailed information, and can improve the classification accuracy of point cloud data. The proposed method was applied to the ModelNet40 and Sydney Urban Objects datasets. In these experiments, FSDCNet outperformed its counterparts, achieving state-of-the-art point cloud classification accuracy. For the ModelNet40 dataset, the overall accuracy (OA) and average accuracy (AA) of FSDCNet in a single view reached 93.8% and 91.2%, respectively, which were superior to those values for many other methods using 6 and 12 views. FSDCNet obtained the best results for 6 and 12 views, achieving 94.6%, 93.3%, 95.3%, and 93.6% in OA and AA metrics, respectively. For the Sydney Urban Objects dataset, FSDCNet achieved an OA and F1 score of 81.2% and 80.1% in a single view, respectively, which were higher than most of the compared methods. In 6 and 12 views, FSDCNet reached an OA of 85.3% and 83.6% and an F1 score of 85.5% and 83.7%, respectively.

On the 3D Track Planning for Electric Power Inspection Based on the Improved Ant Colony Optimization and A∗ Algorithm

At present, multirotor drones are restricted by the control accuracy and cannot position accurately according to the accuracy of point cloud data. Also, track planning in three-dimensional space is much more complicated than that in two-dimensional space, which means that existing track planning methods cannot achieve fast planning. Meanwhile, most existing researches were implemented in quasi-three-dimensional space with the shortest route length as the objective function and omitted environmental impacts. To overcome these, this paper uses the grid method to segment point cloud data of the flying space via ArcGIS software according to the drone’s controlling accuracy. It also extracts the grid coordinate information and maps it to a three-dimensional matrix to build the model accurately. This paper sets the minimal energy consumption as the objective function and builds a track planning model based on the drone’s performance and natural wind constraints. The improved ant colony optimization and A∗ (ACO-A∗) algorithm are utilized to design this algorithm for a faster solution. That is, we use the improved ant colony optimization to quickly find a near-optimal track covering all viewpoints with the minimal energy consumption. The improved A∗ algorithm will be used for local planning for adjacent tracks passing through obstacles. In the designed simulation environment, the simulation results show that, to ensure that the same components are shot, the improved algorithm in this paper can save 62.88% energy compared to that of the Shooting Manual of Drone Inspection Images for Overhead Transmission Lines. Also, it can save 9.33% energy compared to a track with the shortest route length. Besides, the ACO-A∗ algorithm saves 96.6% time than the A∗ algorithm.

Digitisation in facilities management: A literature review and future research directions

Abstract Research into digital technology (DT) in construction practices has gained widespread attention. While the application of different DTs in facility management (FM) has been growing, to date, there is no holistic review of the various DT developments and research into FM. A total of 120 academic journal papers, conference proceedings and other technical notes published on the subject, mainly between 2004 and 2017, were reviewed in this paper. The applications of various major DTs, including 1) building information modelling (BIM), 2) reality capture technology (including 3D laser scanning, point cloud), 3) the Internet of Things (IoT) (including radio frequency identification (RFID) and sensor network technologies) and 4) geographic information system (GIS), were reviewed and scrutinised. The review identified a number of possibilities for future research into DT in FM, including, enhancing the interoperability of data, improving the accuracy of point cloud data for developing as-built models for existing facilities, and generating effective BIM/GIS asset database integration. It is hoped that this review and the future directions highlighted in this paper will assist researchers in identifying the areas where further research efforts are most required and in identifying which future directions would be most helpful for digital FM research.

Estimation of the Density of Forest Based on Point Cloud Data

Based on the Analysis of the characteristics of the point cloud data and the difference compared to other three-dimensional data, explore and study the problem of point cloud data accuracy. From these aspects of the point cloud density, filtering methods and the interpolation method, analysis the elevation accuracy of point cloud data under different categories of terrain. The study method used is in case of different terrain category, different points cloud density and different interpolation methods, using the ground checkpoints of different number, analysis the accuracy of point cloud data. Then use the statistical theory, to estimate the density of Chungking. The final results show that this method can effectively estimate the density of the forest, and has high accuracy.

Estimation of the Density of Forest Based on Point Cloud Data

Based on the Analysis of the characteristics of the point cloud data and the difference compared to other three-dimensional data, explore and study the problem of point cloud data accuracy. From these aspects of the point cloud density, filtering methods and the interpolation method, analysis the elevation accuracy of point cloud data under different categories of terrain. The study method used is in case of different terrain category, different points cloud density and different interpolation methods, using the ground checkpoints of different number, analysis the accuracy of point cloud data. Then use the statistical theory, to estimate the density of Chungking. The final results show that this method can effectively estimate the density of the forest, and has high accuracy.

Determining the limits of unmanned aerial photogrammetry for the evaluation of road runoff

The accuracy of point cloud data in runoff evaluation is a non-resolved question. The increasingly use of unmanned aerial photogrammetric systems shows great opportunities, although they provide less accuracy data than traditional survey systems. This manuscript presents a new method to check the usage limits of point clouds obtained from unmanned aerial photogrammetry in runoff evaluation. The methodology uses D8 algorithm for the runoff estimation. The quality of photogrammetric data is evaluated by comparison with a light detection and ranging scanning dataset that acts as ground truth. The methodology includes point cloud rasterization, with pixel cells ranging from 0.25m to 5m. Five test zones are selected from the road pavement and five from the road slopes. Results show that the runoff evaluation from the unmanned aerial photogrammetric data diverges from the mobile light detection and ranging data as cell size decreases and resolution increases. Studies that require cell sizes below 1m are very influenced by the lower accuracy of the photogrammetric point clouds. However, the unmanned aerial surveys in larger areas with cell sizes above 2m are perfectly affordable for this type of technology, without implication in the quality of runoff results.

Calibration and Accuracy Analysis of a Low-Cost Mapping-Grade Mobile Laser Scanning System

The calibration and accuracy analysis of a novel, low-cost, adaptable mobile laser scanning (MLS) system using a Velodyne HDL-32E laser scanner and an Oxford Technical Solutions Inertial+2 inertial navigation system, is described. First, a static calibration of the laser scanner is discussed. The static calibration is shown to improve the overall relative accuracy of point cloud data from the scanner by approximately 20% over the manufacturer-supplied calibration. Then, the determination of system boresight angles and lever-arm offsets using a planar patch least-squares approach is presented. Finally, the calibrated and boresighted MLS is operated in a backpack mode to acquire multiple data sets in an area that contains dense ground control acquired using static terrestrial laser scanning (TLS) and a high-end, survey-grade MLS. The dense ground control is used to examine several methods of estimating the overall errors of the backpack MLS system. Detailed comparison of the MLS data with the TLS and survey-grade MLS control shows that, despite the system’s low cost, it is able to reliably collect point cloud data with greater than 10-cm three-dimensional root-mean-square error accuracy.

An Improved Algorithm on Point Cloud Optimization for Unmanned Aerial Vehicles

It is hard to keep vertical photography for Unmanned Aerial Vehicles (UAV) and there is large deflection and rotation in the UAV image, which lead to accuracy of point cloud data produced by UAV image is not high. In this paper an algorithm on UAV point cloud optimization based on Patch-based Least Squares Image Matching is proposed. By optimizing the point cloud data with Patch-based Least Squares Image Matching, and test data with images in campus of Northwestern University. The experimental results show that UAV point cloud optimization based on Patch-based Least Squares Image Matching can conspicuously improve the accuracy of the UAV image matching.

Initial validation of point cloud data from a 3D body scanner

PurposeThe purpose of this paper is to validate the accuracy of point cloud data generated from a 3D body scanner.Design/methodology/approachA female dress form was scanned with an X‐ray computed tomography (CT) system and a 3D body scanning system. The point cloud data from four axial slices of the body scan (BS) data were compared with the corresponding axial slices from the CT data. Length and cross‐sectional area measurements of each slice were computed for each scanning technique.FindingsThe point cloud data from the body scanner were accurate to at least 2.0 percent when compared with the CT data. In many cases, the length and area measurements from the two types of scans varied by less than 1.0 percent.Research limitations/implicationsOnly two length measurements and a cross‐sectional area measurement were compared for each axial slice, resulting in a good first attempt of validation of the BS data. Additional methods of comparison should be employed for complete validation of the data. The dress form was scanned only once with each scanning device, so little can be said about the repeatability of the results.Practical implicationsAccuracy of the point cloud data from the 3D body scanner indicates that the main issues for the use of body scanners as anthropometric measurement tools are those of standardization, feature locations, and positioning of the subject.Originality/valueComparisons of point cloud data from a 3D body scanner with CT data had not previously been performed, and these results indicate that the point cloud data are accurate to at least 2.0 percent.