Three-dimensional visualization method and system of power transmission line point cloud

By meshing and blocking the point cloud data of transmission lines, LOD is constructed layer by combining random sampling methods, and deep learning algorithms are used for classification and completion, memory loading and rendering in large-scale point cloud data processing is solved. The problem of low efficiency achieves a more efficient three-dimensional visualization effect.

CN119027615APending Publication Date: 2024-11-26CHINA ENERGY ENG GRP GUANGDONG ELECTRIC POWER DESIGN INST CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411176023.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-26
Publication Date
2024-11-26

AI Technical Summary

Technical Problem

The existing technology encounters the problem that memory cannot be loaded at once when processing large-scale point cloud data, the data quality is insufficient and the web rendering efficiency is low, which affects the user experience.

Method used

By meshing and blocking the point cloud data, LOD is constructed layer by combining random sampling methods, local octree data structure and deep learning algorithm are used for classification and completion, and data loading and rendering efficiency is improved.

Benefits of technology

It solves the memory loading limitations, improves computer stability and rendering efficiency, and improves the quality of point cloud data and three-dimensional visualization effects.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119027615A_ABST
    Figure CN119027615A_ABST
Patent Text Reader

Abstract

The invention discloses a three-dimensional visualization method and system for a point cloud of a power transmission line. The three-dimensional visualization method for the point cloud of the power transmission line comprises the following steps: acquiring point cloud data of the power transmission line; performing grid division and blocking processing on the point cloud data to obtain multi-layer point cloud data; constructing LOD of the point cloud data in a layered manner based on the multi-layer point cloud data and a random sampling method; realizing the three-dimensional visual display of the point cloud of the power transmission line based on the LOD of the point cloud data; according to the method, grid division and blocking processing are carried out on the obtained point cloud data of the power transmission line to obtain the multi-layer point cloud data, the LOD of the point cloud data is constructed in a layered mode based on the multi-layer point cloud data and a random sampling method, when data loading is needed in subsequent three-dimensional visual display, the problem that too large point cloud data cannot be loaded in a memory at a time is solved, and the data loading efficiency is improved. The stability and efficiency of computer operation are greatly improved, the probability of computer rendering collapse is reduced, and then the efficiency of three-dimensional visualization of the power transmission line point cloud can be effectively improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of power transmission line data processing, and in particular to a three-dimensional visualization method and system for power transmission line point cloud. Background Art

[0002] In recent years, with the rapid development of my country's economy, the demand for electricity in all walks of life has continued to increase, the scale of the power grid has continued to expand, and the requirements for the planning, design and maintenance of transmission lines have become increasingly higher. Traditional transmission line design and management mainly rely on two-dimensional maps and manual measurements. This method is not only inefficient, but also difficult to accurately reflect the actual situation of the line. In recent years, with the development of drone technology and lidar technology, the collection and processing of large-scale point cloud data has become possible, which provides a new solution for the design and management of urban-level transmission lines. At the same time, the development of computer graphics and visualization technology has made the three-dimensional visualization of large-scale point cloud data a reality, which can not only improve the efficiency and accuracy of the design, but also intuitively display the actual situation of the line, facilitating decision-making and management. Point cloud data is a set of point data on the surface of a product obtained by a measuring instrument. Each point contains three-dimensional coordinates, and some may also include color information or reflection intensity information. However, due to the large scale of point cloud data, it is very difficult to process. Therefore, it is of great significance to efficiently process large-scale point cloud data and convert it into a format suitable for Web visualization.

[0003] The existing technology mainly processes the point cloud data into blocks, and then converts each block of data into a format suitable for Web visualization, and gradually reduces the resolution of the point cloud data to achieve efficient data loading and display. At the same time, in order to improve the data processing efficiency, the existing technology also uses the octree data structure to store point cloud data. The existing technology has solved the problem of processing large-scale point cloud data to a certain extent, but there are still some problems and shortcomings. First, when processing large-scale point cloud data, the existing technology often encounters the problem that the memory cannot load too much point cloud data at one time. Secondly, due to occlusion, limited sensor resolution and small viewing angle, the original point cloud collected by the scanning device is usually sparse, irregular and incomplete, and the point cloud quality needs to be improved. Finally, the existing technology also has problems in the efficiency of adapting the final data to Web rendering, which may result in a slow display speed on the Web and affect the user experience. Summary of the invention

[0004] The present invention provides a three-dimensional visualization method and medium for power transmission line point cloud to solve the above technical problems and effectively improve the efficiency of three-dimensional visualization of power transmission line point cloud.

[0005] In order to solve the above technical problems, an embodiment of the present invention provides a three-dimensional visualization method of a transmission line point cloud, comprising:

[0006] Obtain point cloud data of transmission lines;

[0007] Perform grid division and block processing on the point cloud data to obtain multi-layer point cloud data;

[0008] Construct LOD of point cloud data hierarchically based on multi-layer point cloud data and random sampling method;

[0009] The three-dimensional visualization display of the transmission line point cloud is realized based on the LOD of the point cloud data.

[0010] The beneficial effects of the present invention are:

[0011] The present invention performs grid division and block processing on the point cloud data of the transmission line to obtain multi-layer point cloud data, and hierarchically constructs the LOD of the point cloud data based on the multi-layer point cloud data and the random sampling method. When data loading is required for subsequent three-dimensional visualization display, the problem that the memory cannot load too much point cloud data at one time is solved, which greatly improves the stability and efficiency of computer operation, reduces the probability of computer rendering crashes, and can effectively improve the efficiency of three-dimensional visualization of transmission line point clouds.

[0012] As a preferred solution, the point cloud data is meshed and divided into blocks, and the multi-layer point cloud data includes:

[0013] Get the coordinates of several towers in the transmission line;

[0014] Mesh the point cloud data to obtain several initial cube meshes;

[0015] Based on the coordinates of each tower, a plurality of first initial cube grids corresponding to each tower are determined respectively;

[0016] Based on the plurality of initial cube grids, a plurality of first initial cube grids corresponding to the same tower are spliced ​​to obtain a plurality of cube grids, thereby completing the grid division of the point cloud data;

[0017] Several cube grids are divided into blocks to obtain multi-layer point cloud data.

[0018] This preferred solution meshes the point cloud data to obtain a number of initial cube grids, and then optimizes the initial cube grids through the tower coordinates. Specifically, the tower coordinates are used to determine in which grid the tower falls. If the same tower falls in different grids, these grids are spliced. In this way, the integrity of the tower point cloud data can be ensured. In addition, this preferred solution is easy to operate and can simply, conveniently and accurately mesh the point cloud data.

[0019] As a preferred solution, the block processing of a plurality of cube grids to obtain multi-layer point cloud data includes:

[0020] Performing a local octree construction on each cube grid to divide each cube grid into a plurality of cube blocks based on a preset local octree level; wherein the preset local octree level is N, and N is a preset positive integer;

[0021] In the Nth level, the number of point clouds of each cube block is compared with the first preset threshold in turn, and a new level of local octree is constructed for the cube blocks with a point cloud number greater than the preset threshold, until the number of point clouds of each cube block in the new level is less than or equal to the preset threshold, thereby obtaining multi-layer point cloud data.

[0022] As a preferred solution, the three-dimensional visualization method of a transmission line point cloud further includes:

[0023] If the sum of the number of point clouds of all cube blocks in the current level is less than the second preset threshold, all cube blocks in the current level are merged into the previous level.

[0024] After the point cloud data is gridded, the preferred solution performs block processing on several cube grids, wherein the point cloud data is hierarchically divided by constructing a local octree for each cube grid to obtain multi-layer point cloud data, and the number of point clouds in each level and each cube block is limited within a preset threshold. This can solve the problem in the prior art that the memory cannot load too much point cloud data at one time, and improve the data loading and processing efficiency.

[0025] As a preferred solution, after completing the grid division of the point cloud data, the method further includes:

[0026] Classify the point cloud data in each cube grid, and obtain the conductor point cloud data, tower point cloud data and transmission line channel environment target point cloud data based on the classification results;

[0027] The conductor point cloud data and the pole tower point cloud data are respectively completed to obtain the completed conductor point cloud data and the completed pole tower point cloud data;

[0028] The completed conductor point cloud data, the completed tower point cloud data and the transmission line channel environment target object point cloud data are fused to update a plurality of cube grids.

[0029] This preferred solution classifies the point cloud data to obtain conductor point cloud data, tower point cloud data and transmission line channel environment target object point cloud data, and improves the quality of the point cloud data by completing, fusing and updating the point cloud data, thereby improving the efficiency of subsequent construction of point cloud detail levels LOD, rendering effects and three-dimensional visualization.

[0030] As a preferred solution, the method of respectively completing the conductor point cloud data and the pole tower point cloud data to obtain the completed conductor point cloud data and the completed pole tower point cloud data includes:

[0031] In the wire point cloud data, obtain the two endpoints of each wire and their endpoint coordinates;

[0032] Determine the initial parameters of the catenary equation based on the endpoint coordinates;

[0033] The objective function is constructed with the goal of minimizing the error between the wire point cloud data and the catenary equation;

[0034] The initial parameters of the catenary equation are optimized based on the objective function to obtain the fitted catenary equation;

[0035] The wire point cloud data is completed based on the fitted catenary equation to obtain the completed wire point cloud data;

[0036] The pole tower point cloud data is completed based on a preset cascade refinement point cloud completion network to obtain the completed pole tower point cloud data.

[0037] This preferred solution constructs an objective function with the goal of minimizing the error between the conductor point cloud data and the catenary equation, optimizes the initial parameters of the catenary equation based on the objective function, obtains the fitted catenary equation, and then completes the conductor point cloud data based on the fitted catenary equation, and completes the tower point cloud data based on a preset cascade refinement point cloud completion network. This allows for the acquisition of denser point cloud data to complete the missing data, improve the quality of the point cloud data and the efficiency of the subsequent construction of the point cloud detail level LOD, so that the subsequent three-dimensional visualization of the transmission line point cloud data has higher visual quality.

[0038] As a preferred solution, after completing the wire point cloud data based on the fitted catenary equation to obtain the completed wire point cloud data, the method further includes:

[0039] The completed wire point cloud data is smoothed to update the completed wire point cloud data.

[0040] This preferred solution performs smoothing on the completed wire point cloud data, which can reduce errors and discontinuities in the data and improve the quality of the point cloud data.

[0041] As a preferred solution, the LOD of point cloud data is constructed hierarchically based on multi-layer point cloud data and random sampling method, including:

[0042] In the first M levels, the first LOD is constructed based on the point cloud data of the target object in the transmission line channel environment; M is a preset positive integer less than N;

[0043] After the M level, starting from the second to last level to the M level, based on the random sampling method, a preset number of point cloud data of each cube block are retained and stored in each level to construct a second LOD.

[0044] As a preferred solution, the LOD based on point cloud data realizes the three-dimensional visualization display of the transmission line point cloud, including:

[0045] Determine a geometric error of each level based on the first LOD and the second LOD to construct 3dtiles of the transmission line point cloud;

[0046] Based on 3dtiles of transmission line point cloud, three-dimensional visualization display of transmission line point cloud is realized.

[0047] The above preferred scheme constructs LOD based on the hierarchy of multi-layer point cloud data. In the first five layers, LOD is constructed for the point cloud data of the transmission line channel environment target objects except the conductor point cloud data and the tower point cloud data. In the latter layers, LOD is constructed for all point clouds, thereby improving the efficiency of layered rendering. Among them, a random sampling method is also used to construct the LOD of point cloud data, which improves the efficiency of point cloud visualization and improves the rendering performance without affecting the three-dimensional visualization visual effect.

[0048] Accordingly, in order to solve the above technical problems, an embodiment of the present invention further provides a three-dimensional visualization system of a transmission line point cloud, comprising: an acquisition module, a processing module, a construction module and a visualization module;

[0049] Wherein, the acquisition module is used to acquire point cloud data of the transmission line;

[0050] The processing module is used to perform grid division and block processing on the point cloud data to obtain multi-layer point cloud data;

[0051] The construction module is used to construct the LOD of point cloud data in layers based on multi-layer point cloud data and random sampling method;

[0052] The visualization module is used to realize three-dimensional visualization display of the transmission line point cloud based on the LOD of the point cloud data. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 : A schematic flow chart of an embodiment of a three-dimensional visualization method of a power transmission line point cloud provided by the present invention;

[0054] Figure 2 : A structural schematic diagram of an embodiment of a three-dimensional visualization system of a transmission line point cloud provided by the present invention. DETAILED DESCRIPTION

[0055] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0056] The existing technology mainly processes the point cloud data into blocks, and then converts each block of data into a format suitable for Web visualization, and gradually reduces the resolution of the point cloud data to achieve efficient data loading and display. At the same time, in order to improve the data processing efficiency, the existing technology also uses the octree data structure to store point cloud data. The existing technology has solved the problem of processing large-scale point cloud data to a certain extent, but there are still some problems and shortcomings. First, when processing large-scale point cloud data, the existing technology often encounters the problem that the memory cannot load too much point cloud data at one time. Secondly, due to occlusion, limited sensor resolution and small viewing angle, the original point cloud collected by the scanning device is usually sparse, irregular and incomplete, and the point cloud quality needs to be improved. Finally, the existing technology also has problems in the efficiency of adapting the final data to Web rendering, which may result in a slow display speed on the Web and affect the user experience.

[0057] Embodiment 1

[0058] In order to solve the above technical problems, the present invention provides a three-dimensional visualization method and medium for transmission line point cloud to solve the above technical problems and effectively improve the efficiency of three-dimensional visualization of transmission line point cloud.

[0059] Please refer to Figure 1 , is a flow chart of an embodiment of a method for three-dimensional visualization of a power transmission line point cloud provided by an embodiment of the present invention. Figure 1 As shown, the three-dimensional visualization method of the transmission line point cloud includes steps 101 to 104, specifically:

[0060] Step 101: Acquire point cloud data of a transmission line;

[0061] Step 102: gridding and block processing are performed on the point cloud data to obtain multi-layer point cloud data;

[0062] Step 103: constructing LOD of point cloud data in layers based on multi-layer point cloud data and random sampling method;

[0063] Step 104: Implement three-dimensional visualization of the transmission line point cloud based on the LOD of the point cloud data.

[0064] The present invention performs grid division and block processing on the point cloud data of the transmission line to obtain multi-layer point cloud data, and hierarchically constructs the LOD of the point cloud data based on the multi-layer point cloud data and the random sampling method. When data loading is required for subsequent three-dimensional visualization display, the problem that the memory cannot load too much point cloud data at one time is solved, which greatly improves the stability and efficiency of computer operation, reduces the probability of computer rendering crashes, and can effectively improve the efficiency of three-dimensional visualization of transmission line point clouds.

[0065] In step 101, the step of obtaining point cloud data of a power transmission line includes:

[0066] Obtaining initial point cloud data of the transmission line;

[0067] The initial point cloud data is preprocessed to obtain point cloud data of the transmission line.

[0068] The present invention preprocesses the initial point cloud data of the transmission line, which can improve the data quality of the point cloud data, so as to improve the efficiency of subsequent grid division, block division and three-dimensional visualization of the point cloud data and improve the effect of three-dimensional visualization.

[0069] In this embodiment, a large-scale point cloud data file of a city-level power transmission line is collected by a UAV equipped with a laser radar to obtain the initial point cloud data of the power transmission line; specifically, the optimal flight altitude and speed are determined, and the UAV is made to fly along the direction of the power transmission line conductor through a route planning algorithm to ensure data coverage and accuracy. The initial point cloud data is obtained by parsing the collected large-scale point cloud data file.

[0070] The specific model of the laser radar is not limited in the present invention, but the Riegl VUX-1LR laser radar is preferably used. The scanning frequency of the radar is 1 MHz, the scanning angle is 360°, and the scanning distance is 1000m, which can meet the needs of large-scale point cloud data collection of urban-level transmission lines.

[0071] In this embodiment, the acquired point cloud data is pre-processed by denoising, filtering, and registration, so as to improve the data quality of the point cloud data.

[0072] Furthermore, in step 102, the point cloud data is meshed and divided into blocks, and the multi-layer point cloud data includes:

[0073] Get the coordinates of several towers in the transmission line;

[0074] Mesh the point cloud data to obtain several cube meshes;

[0075] Based on the coordinates of each tower, a plurality of first initial cube grids corresponding to each tower are determined respectively;

[0076] Based on the plurality of initial cube grids, a plurality of first initial cube grids corresponding to the same tower are spliced ​​to obtain a plurality of cube grids, thereby completing the grid division of the point cloud data;

[0077] Several cube grids are divided into blocks to obtain multi-layer point cloud data.

[0078] The present invention performs grid division on the point cloud data to obtain a plurality of initial cube grids, and then optimizes the initial cube grids through the pole tower coordinates. Specifically, the pole tower coordinates are used to determine in which grid the pole tower falls. If the same pole tower falls in different grids, these grids are spliced, thereby ensuring the integrity of the pole tower point cloud data. Different from the prior art that divides the point cloud data into a preset tree data structure, the present invention is easy to operate and can simply, conveniently and accurately perform grid division on the point cloud data.

[0079] In this embodiment, the point cloud data is divided into a number of initial cube grids according to the plane geographic range, and the point cloud data of each initial cube grid is stored in an external storage octree data structure; the point cloud coordinates are marked as (X, Y, Z), and it is assumed that the side length of the cube grid is set to L. The side length L is determined by calculating the maximum height and the minimum height of the point cloud, as shown in the following formula:

[0080] L=Z max -Z min ;

[0081] Furthermore, the block processing of the plurality of cube grids to obtain multi-layer point cloud data includes:

[0082] Performing a local octree construction on each cube grid to divide each cube grid into a plurality of cube blocks based on a preset local octree level; wherein the preset local octree level is N, and N is a preset positive integer;

[0083] In the Nth level, the number of point clouds of each cube block is compared with the first preset threshold in turn, and a new level of local octree is constructed for the cube blocks with a point cloud number greater than the preset threshold, until the number of point clouds of each cube block in the new level is less than or equal to the preset threshold, thereby obtaining multi-layer point cloud data.

[0084] Furthermore, the three-dimensional visualization method of a transmission line point cloud further includes:

[0085] If the sum of the number of point clouds of all cube blocks in the current level is less than the second preset threshold, all cube blocks in the current level are merged into the previous level.

[0086] After the point cloud data is gridded, the present invention performs block processing on a number of cube grids, wherein the point cloud data is hierarchically divided by constructing a local octree for each cube grid to obtain multi-layer point cloud data, and the number of point clouds in each level and each cube block is limited within a preset threshold. This can solve the problem in the prior art that the memory cannot load too much point cloud data at one time, and improve the data loading and processing efficiency.

[0087] In this embodiment, a local octree is constructed for each cube grid, so as to divide each cube grid into a number of cube blocks based on a preset local octree hierarchy. Specifically, the octree hierarchy is set to 8 layers, that is, N is 8, the top layer is set to 1, and the cube grid is divided into 8 cube blocks according to the size of the cube grid. From the top layer downward, each layer continues to divide the current cube into 8 smaller cube blocks. Thus, starting from the top layer, after 8 equal divisions, each cube grid will be subdivided into 256×256×256 cubes. Furthermore, the number of point clouds of each block is calculated, and then all cube blocks are recursively traversed starting from the 8th layer. If the number of point clouds of the current cube block exceeds 10,000 (the first preset threshold) points, the current cube block is continued to be subdivided into octrees to form block data of the 9th layer, the 10th layer or a new layer; if the total number of point clouds of the 8 cube blocks of the current layer is less than 10,000 (the second preset threshold), the current 8 blocks are merged into the previous layer, and this process is recursively repeated to determine all cube blocks and layers.

[0088] Furthermore, after the point cloud data is meshed, the method further includes:

[0089] Classify the point cloud data in each cube grid, and obtain the conductor point cloud data, tower point cloud data and transmission line channel environment target point cloud data based on the classification results;

[0090] The conductor point cloud data and the pole tower point cloud data are respectively completed to obtain the completed conductor point cloud data and the completed pole tower point cloud data;

[0091] The completed conductor point cloud data, the completed tower point cloud data and the transmission line channel environment target object point cloud data are fused to update a plurality of cube grids.

[0092] The present invention classifies point cloud data to obtain conductor point cloud data, tower point cloud data and transmission line channel environment target object point cloud data, and improves the quality of point cloud data by completing, fusing and updating the point cloud data, thereby improving the efficiency of subsequent construction of point cloud detail level LOD, rendering effect and three-dimensional visualization.

[0093] In this embodiment, the point cloud data of each cube grid can be automatically classified using a deep learning algorithm, but is not limited to, and the characteristics of the transmission line and its channel environment targets are extracted through sample training. Specifically, a training sample is selected from the acquired point cloud data of the transmission line to construct a point cloud classifier, and the point cloud data is classified by the point cloud classifier to divide the point cloud data into categories such as ground points, vegetation points, building points, wires, and poles and towers, that is, the point cloud data is divided into wire point cloud data, pole tower point cloud data, and transmission line channel environment target point cloud data, and then combined with the inspection drone laser radar and the edge computing unit, the transmission line drone inspection three-dimensional scene target recognition is realized, and the wire point cloud data and the pole tower point cloud data are extracted from the point cloud data of the transmission line.

[0094] Furthermore, the said respectively completing the conductor point cloud data and the pole tower point cloud data to obtain the completed conductor point cloud data and the completed pole tower point cloud data comprises:

[0095] In the wire point cloud data, obtain the two endpoints of each wire and their endpoint coordinates;

[0096] Determine the initial parameters of the catenary equation based on the endpoint coordinates;

[0097] The objective function is constructed with the goal of minimizing the error between the wire point cloud data and the catenary equation;

[0098] The initial parameters of the catenary equation are optimized based on the objective function to obtain the fitted catenary equation;

[0099] The wire point cloud data is completed based on the fitted catenary equation to obtain the completed wire point cloud data;

[0100] The pole tower point cloud data is completed based on a preset cascade refinement point cloud completion network to obtain the completed pole tower point cloud data.

[0101] The present invention constructs an objective function with the goal of minimizing the error between the conductor point cloud data and the catenary equation, optimizes the initial parameters of the catenary equation based on the objective function, obtains the fitted catenary equation, and then completes the conductor point cloud data based on the fitted catenary equation, and completes the tower point cloud data based on a preset cascaded refinement point cloud completion network, so that denser point cloud data can be obtained to complete the missing data, improve the quality of the point cloud data and the efficiency of subsequent construction of the point cloud detail level LOD, so that the subsequent three-dimensional visualization of the transmission line point cloud data has higher visual quality.

[0102] In this embodiment, the sag algorithm is used to complete the wire point cloud data, wherein the catenary method is used to calculate the sag, and on this basis, the wire point cloud data is completed. The specific steps are to first obtain the two endpoints of the wire, and then estimate the initial parameters of the catenary according to the coordinates of the endpoints, define an objective function to minimize the error between the wire point cloud data and the catenary equation, use the nonlinear least squares method to optimize the parameters of the catenary equation to obtain the fitted catenary equation, use the fitted catenary equation to generate denser wire point data, combine the points in the fitted curve with the original wire point cloud data, and complete the missing data.

[0103] In this embodiment, the tower point cloud data is completed based on the preset feature fusion cascade refinement point cloud completion network. The preset feature fusion cascade refinement point cloud completion network is a dual-branch network structure, which adopts the strategy from residual point cloud to rough point cloud and then to fine-grained point cloud to complete the tower point cloud data. Among them, the upper branch is used to generate a rough shape with some structural details to generate global features, which can be but not limited to the point completion network (PCN); the PointNet++ algorithm can be but not limited to the lower branch, and the lower branch uses the encoder-decoder structure to generate uniform, dense and fine-grained complete point clouds from global features. The two branches share a global feature to complete the shape completion. The encoder extracts point cloud features of different resolutions and fuses them with the global features. Then the attention mechanism passes the fused point features to the decoder and introduces multiple loss functions to generate a fine point cloud. The present invention completes the tower point cloud data based on a preset feature-fused cascaded refinement point cloud completion network, which not only improves the point cloud quality, but also improves the efficiency of subsequent LOD level construction, so that the method of the present invention has higher visual quality both in terms of global shape and local geometric details.

[0104] Furthermore, after completing the wire point cloud data based on the fitted catenary equation to obtain the completed wire point cloud data, the method further includes:

[0105] The completed wire point cloud data is smoothed to update the completed wire point cloud data.

[0106] The present invention performs smoothing processing on the completed wire point cloud data, which can reduce errors and discontinuities in the data and improve the quality of the point cloud data.

[0107] In this embodiment, one implementation method is to collect point cloud data, perform data preprocessing on the point cloud data, and then grid the point cloud data to obtain a number of cube grids, classify and complete each cube grid, and then perform block processing on the cube grid.

[0108] After completing the block processing and determining all the cube blocks and levels, read the point cloud data again, divide the point cloud data, calculate the spatial coordinates of the current block according to the grid, level, and block, and calculate the coordinates of each point based on the spatial coordinates of the current block. Finally, save each block of point cloud data in a bin file in the "LXYZ" format.

[0109] Further, in step 103, the LOD of the point cloud data is constructed hierarchically based on the multi-layer point cloud data and the random sampling method, including:

[0110] In the first M levels, the first LOD is constructed based on the point cloud data of the target object in the transmission line channel environment; M is a preset positive integer less than N;

[0111] After the M level, starting from the second to last level to the M level, based on the random sampling method, a preset number of point cloud data of each cube block are retained and stored in each level to construct a second LOD.

[0112] In this embodiment, in the hierarchical construction of LOD (level of detail) of point cloud data, since the point cloud data of conductors and pole towers are relatively dense, a simplified model is used in the distance; specifically, LOD is constructed for point clouds such as buildings, vegetation, and ground except for conductors and pole towers in the first five (M) layers, and LOD is constructed for all point clouds in the latter layers. The specific steps are to traverse from the layer above the bottom layer, loop from bottom to top, traverse all blocks of the current layer, obtain the next layer of 8 blocks of data of the current block, use a random sampling method to take less than 10,000 points, and store them in the current block, and recursively perform this operation to construct LOD.

[0113] Further, in step 104, the LOD based on the point cloud data realizes the three-dimensional visualization display of the transmission line point cloud, including:

[0114] Determine a geometric error of each level based on the first LOD and the second LOD to construct 3dtiles of the transmission line point cloud;

[0115] Based on 3dtiles of transmission line point cloud, three-dimensional visualization display of transmission line point cloud is realized.

[0116] In this embodiment, the geometric error of each level is determined based on the first LOD and the second LOD to construct the 3dtiles of the transmission line point cloud. Specifically, the geometricError of the first layer is first determined based on the first LOD and the second LOD. The first layer is a grid divided according to the plane geographical range. The spatial size of the grid is calculated, and the side length / 256 of the cube grid is taken as the geometricError of each node in the first layer. Then, the octree in each cube grid is divided according to geometricError / 2 to determine the geometric error of each level. Based on the geometric error of each level, the corresponding pnts files are generated for the cube block data of all levels, and then the tileset.json file is uniformly constructed to recursively construct the hierarchical structure of 3dtiles.

[0117] When the above scheme constructs LOD levels based on the levels of multi-layer point cloud data, the first five layers construct LOD for the point cloud data of the transmission line channel environment target objects except the conductor point cloud data and the tower point cloud data, and the latter layers construct LOD for all point clouds, thereby improving the efficiency of layered rendering. Among them, a random sampling method is also used to construct the LOD of point cloud data, and a transmission line point cloud pyramid model is constructed, which improves the efficiency of point cloud visualization and improves the rendering performance without affecting the three-dimensional visualization visual effect.

[0118] Accordingly, in order to solve the above technical problems, the embodiment of the present invention also provides a three-dimensional visualization system of a transmission line point cloud. Figure 2 , which is a structural schematic diagram of an embodiment of a three-dimensional visualization system of a transmission line point cloud provided by the present invention.

[0119] like Figure 2 As shown, the three-dimensional visualization system 20 of the transmission line point cloud includes: an acquisition module 201, a processing module 202, a construction module 203 and a visualization module 204;

[0120] Wherein, the acquisition module 201 is used to acquire point cloud data of the transmission line;

[0121] The processing module 202 is used to perform grid division and block processing on the point cloud data to obtain multi-layer point cloud data;

[0122] The construction module 203 is used to construct the LOD of the point cloud data in layers based on the multi-layer point cloud data and the random sampling method;

[0123] The visualization module 204 is used to realize three-dimensional visualization display of the transmission line point cloud based on the LOD of the point cloud data.

[0124] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0125] The present invention performs grid division on large-scale point cloud data of power transmission lines, stores the data in an external memory octree data structure, and constructs a local octree for each cube grid to divide the data into blocks, thereby solving the problem that the memory cannot load too much point cloud data at one time, improving data loading and processing efficiency, greatly improving the stability and efficiency of computer operation, and reducing the probability of computer rendering crashes.

[0126] The present invention adopts deep learning algorithm and sag algorithm to automatically classify and complete point cloud data, improves the data quality of conductor point cloud data and pole tower point cloud data, and extracts conductors and pole towers to improve the efficiency of subsequent LOD construction.

[0127] The present invention combines the point cloud extraction results of the conductors and towers and adopts the LOD algorithm based on random sampling to construct a point cloud pyramid model of the transmission line, thereby improving the efficiency of three-dimensional visualization of the point cloud and improving the rendering performance without affecting the visual effect.

[0128] The specific embodiments described above further illustrate the purpose, technical solutions and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. It is particularly pointed out that for those skilled in the art, any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the scope of protection of the present invention.

Claims

1. A three-dimensional visualization method for a transmission line point cloud, characterized in that: include: Obtain point cloud data of transmission lines; Perform grid division and block processing on the point cloud data to obtain multi-layer point cloud data; Construct LOD of point cloud data hierarchically based on multi-layer point cloud data and random sampling method; The three-dimensional visualization display of the transmission line point cloud is realized based on the LOD of the point cloud data.

2. A three-dimensional visualization method for a power transmission line point cloud according to claim 1, characterized in that: The point cloud data is meshed and divided into blocks, and the multi-layer point cloud data includes: Get the coordinates of several towers in the transmission line; Mesh the point cloud data to obtain several initial cube meshes; Based on the coordinates of each tower, a plurality of first initial cube grids corresponding to each tower are determined respectively; Based on the plurality of initial cube grids, a plurality of first initial cube grids corresponding to the same tower are spliced ​​to obtain a plurality of cube grids, thereby completing the grid division of the point cloud data; Several cube grids are divided into blocks to obtain multi-layer point cloud data.

3. A three-dimensional visualization method for a power transmission line point cloud according to claim 2, characterized in that: The block processing of a plurality of cube grids to obtain multi-layer point cloud data includes: Performing a local octree construction on each cube grid to divide each cube grid into a plurality of cube blocks based on a preset local octree level; wherein the preset local octree level is N, and N is a preset positive integer; In the Nth level, the number of point clouds of each cube block is compared with the first preset threshold in turn, and a new level of local octree is constructed for the cube blocks with a point cloud number greater than the preset threshold, until the number of point clouds of each cube block in the new level is less than or equal to the preset threshold, thereby obtaining multi-layer point cloud data.

4. A three-dimensional visualization method for a power transmission line point cloud according to claim 3, characterized in that: Also includes: If the sum of the number of point clouds of all cube blocks in the current level is less than the second preset threshold, all cube blocks in the current level are merged into the previous level.

5. The three-dimensional visualization method of a power transmission line point cloud according to claim 2, characterized in that: After the point cloud data is meshed, the method further includes: Classify the point cloud data in each cube grid, and obtain the conductor point cloud data, tower point cloud data and transmission line channel environment target point cloud data based on the classification results; The conductor point cloud data and the pole tower point cloud data are respectively completed to obtain the completed conductor point cloud data and the completed pole tower point cloud data; The completed conductor point cloud data, the completed tower point cloud data and the transmission line channel environment target object point cloud data are fused to update a plurality of cube grids.

6. A three-dimensional visualization method for a power transmission line point cloud according to claim 5, characterized in that: The method of respectively completing the conductor point cloud data and the pole tower point cloud data to obtain the completed conductor point cloud data and the completed pole tower point cloud data comprises: In the wire point cloud data, obtain the two endpoints of each wire and their endpoint coordinates; Determine the initial parameters of the catenary equation based on the endpoint coordinates; The objective function is constructed with the goal of minimizing the error between the wire point cloud data and the catenary equation; The initial parameters of the catenary equation are optimized based on the objective function to obtain the fitted catenary equation; The wire point cloud data is completed based on the fitted catenary equation to obtain the completed wire point cloud data; The pole tower point cloud data is completed based on a preset cascade refinement point cloud completion network to obtain the completed pole tower point cloud data.

7. A three-dimensional visualization method for a power transmission line point cloud according to claim 6, characterized in that: After completing the wire point cloud data based on the fitted catenary equation to obtain the completed wire point cloud data, the method further includes: The completed wire point cloud data is smoothed to update the completed wire point cloud data.

8. The three-dimensional visualization method of a power transmission line point cloud according to claim 5, characterized in that: The method of constructing LOD of point cloud data in layers based on multi-layer point cloud data and random sampling method includes: In the first M levels, the first LOD is constructed based on the point cloud data of the target object in the transmission line channel environment; M is a preset positive integer less than N; After the M level, starting from the second to last level to the M level, based on the random sampling method, a preset number of point cloud data of each cube block are retained and stored in each level to construct a second LOD.

9. A three-dimensional visualization method for a power transmission line point cloud according to claim 8, characterized in that: The LOD based on point cloud data realizes the three-dimensional visualization display of the transmission line point cloud, including: Determine a geometric error of each level based on the first LOD and the second LOD to construct 3dtiles of the transmission line point cloud; Based on 3dtiles of transmission line point cloud, three-dimensional visualization display of transmission line point cloud is realized.

10. A three-dimensional visualization system for transmission line point clouds, characterized in that: include: Acquisition module, processing module, construction module and visualization module; Wherein, the acquisition module is used to acquire point cloud data of the transmission line; The processing module is used to perform grid division and block processing on the point cloud data to obtain multi-layer point cloud data; The construction module is used to construct the LOD of point cloud data in layers based on multi-layer point cloud data and random sampling method; The visualization module is used to realize three-dimensional visualization display of the transmission line point cloud based on the LOD of the point cloud data.