A template matching method and system for tower point cloud

Abstract

This invention discloses a template matching method and system for pole point clouds. The template matching method includes: scanning a target pole area using a 3D LiDAR to extract pole point cloud data; selecting a corresponding template type from a preset pole template database based on the point cloud shape; extracting pole outline feature points from the pole point cloud data according to the template type; matching the pole outline feature points with model feature points of the corresponding template type in the preset pole template database to obtain the spatial position of the successfully matched model feature points; and scaling the remaining model feature points of the corresponding template type proportionally based on the spatial position of the successfully matched model feature points to obtain the pole model corresponding to the pole point cloud data. This invention solves the problem of excessive manual workload in existing technologies.

Description

Technical Field

[0001] This invention relates to the field of power grid technology, and in particular to a template matching method and system for tower point clouds. Background Technology

[0002] With the continuous strengthening of power grid construction, the scale of the power grid is also expanding rapidly. To adapt to the expansion of the power grid, 3D digital power grid technology is becoming increasingly important. In 3D digital power grid technology, power poles are an important structural component of transmission lines, directly affecting the reliability and safety of the transmission lines. Therefore, 3D digital power pole modeling technology has become an indispensable part of 3D digital power grid technology. Traditional 3D digital pole modeling technologies mainly include modeling methods based on 2D drawings, oblique photogrammetry, and 3D laser scanning. Among them, the 2D drawing-based method requires the completion of 2D design drawings and the use of 3D software to model the 2D design drawings. The model construction is complex, costly, and time-consuming. The oblique photogrammetry-based method requires the use of aerial triangulation to calculate the 3D model. The model data volume is large, the calculation process is time-consuming, the accuracy is poor, and it is easily affected by the environment.

[0003] Compared to the two modeling methods mentioned above, the pole modeling method based on 3D laser scanning, while also facing problems such as large amounts of repetitive data and complex operations, boasts high point cloud accuracy and can address the data volume issue through basic model construction, thus possessing significant development potential. Existing pole modeling methods based on 3D laser scanning mainly fall into two categories: 1. Layered pole point cloud modeling; 2. Modeling using a pole model component library. The layered pole point cloud modeling method specifically uses layered pole point clouds, calculating key pole points using parameters such as the number of points in each layer, and then establishing the topological relationships between these key points to build the pole model. This method primarily determines key points and topological relationships through feature extraction; however, due to variations in the acquisition density of pole point clouds and uneven point counts in each layer, errors in the resulting topological relationships may occur. The pole model component library modeling method involves constructing a pole model component library. Since the number of pole models is limited, a layered component library is first built, and then the components are connected to construct the final pole model. Specifically, the model is constructed by selecting the tower head and tower foot components and then setting the tower parameters. For example, the transmission line tower based on a full-element model library of laser point clouds and its implementation method provided by patent CN110136260A constructs a component model library through laser point clouds, separates the tower head and tower foot, and then records the connection point information of the tower head and tower foot.

[0004] However, the creation of a hierarchical component library requires tasks such as setting connection point topology relationships and tower parameters, resulting in a large amount of manual work. Moreover, the above method requires setting similar adjustment parameters or performing various repetitive operations when dealing with the same type of tower point cloud data, further increasing the workload. Summary of the Invention

[0005] This invention provides a template matching scheme for tower point clouds, aiming to solve the problem of large manual workload caused by various tasks and repetitive operations in the prior art.

[0006] To address the aforementioned problems, according to a first aspect of the present invention, a template matching method for tower point clouds is proposed, comprising:

[0007] The target tower area was scanned using a 3D LiDAR scanner to extract tower point cloud data.

[0008] Based on the point cloud shape of the tower point cloud data, select the corresponding template type from the preset tower template database;

[0009] Based on the template type, extract tower outline feature points from tower point cloud data;

[0010] Match the tower outline feature points with the model feature points of the corresponding template type in the preset tower template database to obtain the spatial location of the successfully matched model feature points;

[0011] Based on the spatial location of the successfully matched model feature points, the remaining model feature points of the corresponding template type are scaled proportionally to obtain the tower model corresponding to the tower point cloud data.

[0012] Preferably, in the above-mentioned template matching method for tower point clouds, the step of using a 3D lidar to scan the target tower area and extract the tower point cloud data includes:

[0013] The target power line corridor was scanned using a 3D LiDAR scanner to obtain point cloud data of the target power line corridor.

[0014] The point cloud data of the target power line corridor is denoised and classified to obtain point cloud data of various power line types;

[0015] Pole and tower point cloud data were extracted from point cloud data of various power line types.

[0016] Preferably, in the above-mentioned template matching method for tower point clouds, the step of selecting the corresponding template type from a preset tower template database based on the point cloud shape of the tower point cloud data includes:

[0017] Select the pole image shape that corresponds to the point cloud shape from the preset pole template database;

[0018] Determine the tower model corresponding to the shape of the tower image;

[0019] Search the preset pole template database for the template type corresponding to the pole model;

[0020] Extract the tower template corresponding to the template type.

[0021] Preferably, the above-mentioned pole point cloud template matching method further includes, before the step of selecting the corresponding template type from a preset pole template database based on the point cloud shape of the pole point cloud data:

[0022] Search the preset tower template database for the corresponding tower template type, and use the inner and outer contour feature points of the tower template as the first type of model feature points;

[0023] Based on the collinear relationship between each node in the tower template and the feature points of the first type of model, other types of tower nodes in the tower template are obtained;

[0024] Connect the feature points of the first type of model with the nodes of other types of towers to obtain the tower element.

[0025] Preferably, in the above-mentioned template matching method for tower point clouds, the step of extracting tower contour feature points from the tower point cloud data according to the template type includes:

[0026] Determine the tower template corresponding to the template type;

[0027] Extract the undulation structure of the first type of model feature points in the tower template, and the undulation structure of the tower outline in the tower point cloud data; wherein, the first type of model feature points are the inner and outer contour feature points of the tower template;

[0028] Position matching is performed on the undulation structure of the feature points of the first type of model and the concave and convex points in the undulation structure of the tower profile.

[0029] If the positions of the concave and convex points in the undulating structure are successfully matched, the concave and convex points in the tower point cloud data are extracted as tower contour feature points.

[0030] Preferably, in the above-mentioned template matching method for tower point clouds, the step of matching tower contour feature points with model feature points of the corresponding template type in a preset tower template database to obtain the spatial location of the successfully matched model feature points includes:

[0031] Select the pole template corresponding to the template type from the preset pole template database;

[0032] Feature matching is performed between the feature points of the tower outline and the feature points of the first type of model in the tower template;

[0033] When the feature points of the tower outline are successfully matched with the feature points of the first type of model, the spatial position of the tower outline feature points is taken as the spatial position of the matched feature points of the first type of model.

[0034] Based on the distance ratio between other types of tower nodes in the tower template and the matching first type of model feature points, the spatial positions of other types of tower nodes are obtained.

[0035] Preferably, in the above-mentioned template matching method for tower point clouds, the step of performing feature matching between the tower contour feature points and the first type of model feature points in the tower template includes:

[0036] Based on the fluctuation structure of the feature points of the first type of model, the feature points of the first type of model are divided into coarse feature points and fine feature points;

[0037] Feature matching is performed on the tower outline feature points and coarse feature points to obtain the successfully matched coarse feature points;

[0038] Based on the positional relationship between the successfully matched coarse feature points and fine feature points, as well as the positional relationship between the fine feature points, the fine feature points that successfully match the tower outline feature points are obtained.

[0039] Preferably, in the above-mentioned template matching method for tower point clouds, the step of scaling the remaining model feature points of the corresponding template type proportionally according to the spatial location of the successfully matched model feature points to obtain the tower model corresponding to the tower point cloud data includes:

[0040] Based on the feature matching results between the tower outline feature points and the first type of model feature points, the first type of model feature points are translated to the spatial positions of the corresponding tower outline feature points.

[0041] Based on the point-to-point ratio between other types of tower nodes and the feature points of the first type of model, the other types of tower nodes are scaled to obtain the tower model corresponding to the tower point cloud data.

[0042] According to a second aspect of the present invention, the present invention also provides a template matching system for tower point clouds, comprising:

[0043] The pole and tower point cloud extraction module is used to scan the target pole and tower area with a 3D LiDAR and extract the pole and tower point cloud data.

[0044] The template type selection module is used to select the corresponding template type from the preset tower template database based on the point cloud shape of the tower point cloud data;

[0045] The contour feature extraction module is used to extract tower contour feature points from tower point cloud data based on template type.

[0046] The model feature matching module is used to match tower outline feature points with model feature points of the corresponding template type in the preset tower template database, and obtain the spatial location of the successfully matched model feature points.

[0047] The model feature scaling module scales the remaining model feature points of the corresponding template type proportionally according to the spatial location of the successfully matched model feature points, thereby obtaining the tower model corresponding to the tower point cloud data.

[0048] According to a third aspect of the present invention, the present invention also provides a template matching system for tower point clouds, comprising:

[0049] The present invention includes a memory, a processor, and a template matching program for pole point clouds stored in the memory and executable on the processor. When the template matching program for pole point clouds is executed by the processor, it implements the steps of the template matching method for pole point clouds provided by any of the above technical solutions.

[0050] In summary, the pole point cloud template matching scheme provided by this invention extracts pole point cloud data by scanning the target pole area with a 3D LiDAR scanner. Then, based on the point cloud shape, a corresponding template type is selected from a preset pole template database. Next, pole outline feature points are extracted from the pole point cloud data. This allows for automatic matching of these feature points with model feature points of the corresponding template type in the preset template database, resulting in the spatial position of the successfully matched model feature points. These model feature points then share the same spatial position as the pole outline feature points. By scaling the remaining model feature points proportionally, a complete pole model corresponding to the pole point cloud data can be obtained. This allows for matching the corresponding template type to each type of pole point cloud data, thus achieving automatic pole modeling. In conclusion, this method solves the problems of repetitive work and high time consumption in existing model building processes for similar pole point cloud data, enabling efficient construction of usable models corresponding to pole point cloud data and saving labor costs. In addition, it solves the problems of complex operation and too many parameters in the existing tower modeling process. By setting up the tower template database and matching the corresponding template type for targeted modeling, the manual operation cost can be greatly reduced, and the problem of large workload caused by various operation tasks and repetitive operations in the existing technology can be solved. Attached Figure Description

[0051] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0052] Figure 1 This is a flowchart illustrating a template matching method for tower point clouds provided in an embodiment of the present invention;

[0053] Figure 2 yes Figure 1 The illustrated embodiment provides a flowchart of a method for extracting point cloud data from poles and towers.

[0054] Figure 3 yes Figure 1 The illustrated embodiment provides a flowchart of a template type selection method;

[0055] Figure 4 yes Figure 1 The illustrated embodiment provides a flowchart of a method for establishing a preset tower template database;

[0056] Figure 5 yes Figure 1 The illustrated embodiment provides a flowchart of a method for extracting feature points of a tower profile.

[0057] Figure 6 yes Figure 1 The illustrated embodiment provides a flowchart of a method for matching the spatial location of model feature points;

[0058] Figure 7 yes Figure 6 The illustrated embodiment provides a flowchart of a feature point feature matching method;

[0059] Figure 8 yes Figure 1 The illustrated embodiment provides a flowchart of a method for proportional scaling of model feature points;

[0060] Figure 9 This is a schematic diagram of the structure of a tower model of a wine glass tower provided in an embodiment of the present invention;

[0061] Figure 10 This is a flowchart illustrating the first template matching method for tower point clouds provided in this embodiment of the invention.

[0062] Figure 11 This is a flowchart illustrating the second method for template matching of tower point clouds provided in this embodiment of the invention.

[0063] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0064] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0065] The main technical problem solved by the embodiments of the present invention is:

[0066] There are two main existing methods for pole modeling based on 3D laser scanning: 1. Modeling through layered pole point clouds; 2. Modeling by constructing a pole model component library. The latter method involves building a pole model component library. Due to the limited number of pole models, a layered component library is first constructed, and then the components are connected to create the final pole model. Building a layered component library requires setting connection point topology relationships and pole parameters, resulting in a significant manual workload. Furthermore, when dealing with similar pole point cloud data, the above methods require setting similar adjustment parameters or performing various repetitive operations, further increasing the workload.

[0067] To address the aforementioned issues, the following embodiments of the present invention provide a template matching scheme for pole point clouds. By using a preset pole template database, a corresponding template type is selected from the database. Then, pole outline feature points are extracted from the pole point cloud data. This automatically matches the pole outline feature points with the model feature points of the corresponding template type in the data block, obtaining the spatial location of the successfully matched model feature points. The remaining model feature points corresponding to the template type are then scaled proportionally to obtain the complete pole model corresponding to the pole point cloud data. Thus, each type of pole point cloud data is matched with a corresponding template type, achieving automatic modeling and solving the problem of excessive workload caused by various operational tasks and repetitive operations in existing technologies.

[0068] To achieve the above objectives, see [link to relevant documentation]. Figure 1 , Figure 1 This is a flowchart illustrating a template matching method for pole point clouds provided in an embodiment of the present invention. Figure 1 As shown, the template matching method for the tower point cloud includes:

[0069] S110: Use a 3D LiDAR scanner to scan the target tower area and extract tower point cloud data. By scanning the target tower area and extracting the tower point cloud data, the point cloud data can be compared and matched with a tower template of a specific template type to determine the spatial location of the template's feature points. The target tower area includes areas such as power line corridors, specifically containing data on conductors, towers, ground, and vegetation.

[0070] Specifically, as a preferred embodiment, such as Figure 2 As shown, the steps described above for using a 3D lidar to scan the target tower area and extract the tower point cloud data include:

[0071] S111: Use a 3D LiDAR scanner to scan the target power line corridor and obtain point cloud data of the target power line corridor. The point cloud data collected by scanning the target power line corridor with a 3D LiDAR scanner mainly includes point cloud data of conductors, towers, ground, and vegetation.

[0072] S112: Denoise and classify the point cloud data of the target power line corridor to obtain point cloud data of various power line types.

[0073] S113: Extract pole point cloud data from point cloud data of various power line types.

[0074] Specifically, point cloud data of various power line types describes the absolute spatial location of objects on Earth. These points are distributed throughout the scanned area and contain various useful ground feature information, such as ground points, pole points, conductor points, and vegetation points, as well as useless point data such as noise points. Therefore, the point cloud data needs to be denoised and classified to filter out noise and then classify usable point cloud data. Classification methods for point cloud data of the target power line corridor include manual classification, point cloud clustering, or artificial intelligence classification. After classification, the initial classification results are optimized and corrected to obtain the required pole point cloud data containing ground points, pole points, and conductor points.

[0075] Figure 1 The template matching method for tower point clouds provided in the illustrated embodiment further includes, after extracting the tower point cloud data:

[0076] S120: Based on the point cloud shape of the tower point cloud data, select the corresponding template type from the preset tower template database. In this embodiment of the invention, for specific tower point cloud data, the corresponding template type is selected from the preset tower template database according to the point cloud shape. Specifically, the feature points corresponding to the template type are selected. This allows for matching of point clouds of different tower shapes, thereby enabling the modeling of towers of different shapes.

[0077] Specifically, as a preferred embodiment, such as Figure 3 As shown, in the above-mentioned pole point cloud template matching method, step S120: selecting the corresponding template type from the preset pole template database based on the point cloud shape of the pole point cloud data includes:

[0078] S121: Select the pole image shape corresponding to the point cloud shape from the preset pole template database.

[0079] S122: Determine the tower model corresponding to the shape of the tower image.

[0080] S123: Search for the template type corresponding to the tower model from the preset tower template database.

[0081] S124: Extract the tower template corresponding to the template type.

[0082] In the technical solution provided by this invention, the preset tower template database includes various tower models, with common tower models constituting tower templates. Each tower model includes tower nodes and tower units. Tower nodes include key points and their positions, while units are the lines connecting two tower nodes. Tower nodes are divided into four categories, the first being model feature points. Thus, after determining the template type corresponding to the point cloud shape of the tower point cloud data, a tower template corresponding to that template type can be selected, thereby accurately extracting the spatial positions of tower nodes in the tower template and automatically and accurately depicting the tower models corresponding to various tower point cloud data. Furthermore, the selection criterion for the template type depends on the tower image shape corresponding to the point cloud shape in the preset tower template database. For tower models corresponding to the same tower image shape, the internal algorithm extracts the corresponding tower feature points based on the template type for subsequent matching calculations. Selecting an unmatched tower template may result in tower template matching failure.

[0083] In addition, before selecting the template type corresponding to the pole point cloud data from the preset pole template database, it is necessary to establish the preset pole template database and complete the modeling of each pole model, pole template, pole node and pole unit within it.

[0084] Specifically, as a preferred embodiment, such as Figure 4 As shown, the above-mentioned pole point cloud template matching method, before step S120: selecting the corresponding template type from the preset pole template database based on the point cloud shape of the pole point cloud data, further includes:

[0085] S210: Search for the pole template corresponding to the template type from the preset pole template database, and take the inner and outer contour feature points of the pole template as the first type of model feature points.

[0086] S220: Based on the collinear relationship between each node in the tower template and the feature points of the first type of model, other types of tower nodes in the tower template are obtained.

[0087] S230: Connect the feature points of the first type of model with the nodes of other types of towers to obtain the tower element.

[0088] In the technical solution provided by this invention, the tower template is a common tower model among specific template types. Typically, a tower model consists of nodes and elements; common tower models among template types constitute the tower template. Specifically, as follows... Figure 9 As shown, Figure 9This is a structural schematic diagram of a tower model for a wine glass-shaped tower. Based on this tower model, it can be seen that the nodes in the tower model are divided into a first category of model feature points and other categories of tower nodes. These other categories of tower nodes include second, third, and auxiliary material nodes; among them,

[0089] The first type of model feature points are the key nodes of the tower model, such as... Figure 9 As shown in node 3, these mainly represent the feature points of the inner and outer contours of the tower template, including the "inflection points" of the inner and outer contours or the gradient change points of the tower contour.

[0090] The feature points of the second type of model are generated by the collinear relationship between two first type nodes;

[0091] The third type of model feature points are formed by collinearity between the first type of model feature points and the second type of model feature points, or entirely by collinearity between the second type of model feature points (that is, the first type of model feature points of the tower model are represented by the absolute position of the points, and the second and third type of model feature points are generated by point pairs according to the distance ratio).

[0092] The auxiliary material nodes are constructed in the manner of the third type of model feature points (their constituent units are not subject to force, unlike the other nodes).

[0093] The unit of the pole model is composed of nodes (including the feature points of the first type of model mentioned above and other types of pole nodes) connected together.

[0094] in addition, Figure 1 The template matching method for tower point clouds provided in the illustrated embodiment, after selecting the corresponding template type based on the point cloud shape of the tower point cloud data, further includes the following steps:

[0095] S130: Extract tower contour feature points from the tower point cloud data according to the template type. The template type corresponds one-to-one with the tower type in the tower point cloud data; among all template types, common tower models constitute the tower template. The extraction of tower contour feature points is related to the first type of model feature points in the tower template.

[0096] Specifically, as a preferred option, such as Figure 5 As shown, in the template matching method for tower point clouds described above, step S130: extracting tower contour feature points from the tower point cloud data according to the template type includes:

[0097] S131: Determine the tower template corresponding to the template type.

[0098] S132: Extract the undulation structure of the first type of model feature points in the tower template, and the undulation structure of the tower outline in the tower point cloud data; wherein, the first type of model feature points are the inner and outer contour feature points of the tower template.

[0099] S133: Perform position matching between the concave and convex points of the undulating structure of the feature points of the first type of model and the concave and convex points of the undulating structure of the tower profile.

[0100] S134: If the positions of the concave and convex points in the undulating structure are successfully matched, then the concave and convex points in the tower point cloud data are extracted as tower contour feature points.

[0101] In the technical solution provided by this invention, after determining the template type corresponding to the tower point cloud data, the tower contour feature points are extracted from the tower point cloud data according to the positions of the tower contour feature points provided by the template type. Specifically, the tower template corresponding to the template type is determined, and the undulation structure of the first type of model feature points and the undulation structure of the tower contour in the tower template are extracted. The positions of the convex and concave points in the undulation structure of the first type of model feature points and the undulation structure of the tower contour are matched. After successful matching, the convex and concave points in the tower point cloud data are extracted as tower contour feature points. This allows for accurate spatial positioning of the tower contour feature points. Specifically, the model feature points are determined by the first type of feature points in the template. The extraction algorithm also performs position matching between the undulation relationship of the first type of feature points in the tower template and the undulation relationship of the tower outer contour in the tower point cloud data. Point-to-point position comparison matching is performed on both sides according to whether they are convex or concave points, thereby accurately extracting the tower contour feature points from the point cloud data.

[0102] Figure 1 The template matching method for tower point clouds provided in the illustrated embodiment further includes the following steps after extracting tower contour feature points from the tower point cloud data:

[0103] S140: Match the tower outline feature points with the corresponding model feature points in the preset tower template database to obtain the spatial positions of the successfully matched model feature points. By matching the tower outline feature points with the model feature points in the preset tower template database, the spatial positions of the model feature points corresponding to the tower outline feature points can be determined. This allows for the description of the outline of different types of tower point cloud data, obtaining the spatial positions of each node in the corresponding model, and thus accurately describing the spatial structure of the tower point cloud data.

[0104] Specifically, as a preferred embodiment, such as Figure 6 As shown, in the above template matching method for tower point clouds, step S140: matching tower contour feature points with model feature points of the corresponding template type in the preset tower template database to obtain the spatial location of the successfully matched model feature points includes:

[0105] S141: Select the pole template corresponding to the template type from the preset pole template database.

[0106] S142: Perform feature matching between the feature points of the tower outline and the feature points of the first type of model in the tower template.

[0107] S143: When the feature points of the tower outline are successfully matched with the feature points of the first type of model, the spatial position of the tower outline feature points is taken as the spatial position of the matched feature points of the first type of model.

[0108] S144: Based on the distance ratio between other types of tower nodes in the tower template and the matching first type of model feature points, obtain the spatial positions of other types of tower nodes.

[0109] In the technical solution provided by this invention, the matching of tower templates mainly involves matching the first type of model feature points of tower templates in the database, while the spatial positions of other types of tower nodes in the tower templates are obtained according to proportional relationships. The preset tower template database contains tower models of various template types, and different types of tower templates correspond to different model feature points. Based on the settings, the corresponding first type of model feature points (i.e., model feature points specifically extracted from the current tower point cloud data corresponding to the tower template type) are automatically extracted. These first type of model feature points are then matched with other types of tower structures in the tower template, thereby obtaining the accurate spatial positions of the first type of model feature points and other types of tower nodes.

[0110] As a preferred embodiment, such as Figure 7 As shown, in the above template matching method for tower point clouds, step S142: the step of matching the tower contour feature points with the first type of model feature points in the tower template includes:

[0111] S1421: Based on the fluctuation structure of the feature points of the first type of model, the feature points of the first type of model are divided into coarse feature points and fine feature points.

[0112] S1422: Perform feature matching between the tower outline feature points and coarse feature points to obtain successfully matched coarse feature points.

[0113] S1423: Based on the positional relationship between the successfully matched coarse feature points and fine feature points, as well as the positional relationship between the fine feature points, obtain the fine feature points that are successfully matched with the tower outline feature points.

[0114] In the technical solution provided by this invention, the first type of model feature points are divided into coarse model feature points and fine model feature points during the matching process; see details below. Figure 9In the wine glass tower model shown, the undulations in the bottom structural frame 1 are obvious and easy to extract, and the corresponding feature points are coarse feature points. However, in the top structural frame 2, the feature points are not easy to extract and need to be determined by other feature points; these are fine feature points. The feature information of the coarse feature points is significantly easier to extract, while the feature information of the fine feature points depends on the results of coarse matching points. After obtaining the coarse and fine feature points, feature matching is performed between the tower outline feature points and the coarse feature points to obtain the successfully matched coarse feature points. Then, based on the positional relationship (e.g., coordinate distance) between the coarse and fine feature points, and the positional relationship between the fine feature points themselves, the positions of the fine feature points corresponding to the aforementioned tower outline feature points can be obtained.

[0115] Figure 1 The template matching method for tower point clouds provided in the illustrated embodiment further includes, after obtaining the spatial location of the successfully matched model feature points:

[0116] S150: Based on the spatial location of the successfully matched model feature points, scale the remaining model feature points of the corresponding template type proportionally to obtain the tower model corresponding to the tower point cloud data.

[0117] After obtaining the spatial location of the successfully matched model feature points, the remaining model feature points (such as the other types of tower nodes mentioned above) are scaled proportionally to quickly match the spatial location of the remaining model feature points in the tower point cloud data. In this way, after obtaining the spatial locations of the first type of model feature points of the tower template and the other types of tower nodes mentioned above in the tower point cloud data, the tower model corresponding to the tower point cloud data can be accurately obtained.

[0118] Specifically, as a preferred embodiment, such as Figure 8 As shown, the steps described above, which involve scaling the remaining model feature points of the corresponding template type proportionally based on the spatial location of the successfully matched model feature points to obtain the tower model corresponding to the tower point cloud data, specifically include:

[0119] S151: Based on the feature matching results between the tower outline feature points and the first type of model feature points, translate the first type of model feature points to the spatial position of the corresponding tower outline feature points.

[0120] S152: Scale the other types of tower nodes according to the point-to-point ratio between the other types of tower nodes and the feature points of the first type of model to obtain the tower model corresponding to the tower point cloud data.

[0121] The technical solution provided in this invention, after determining the spatial positions of the first type of model feature points in the successfully matched model based on the feature point correspondence results, directly translates the first type of model feature points to the corresponding tower point cloud positions. The spatial positions of the remaining second, third, and auxiliary nodes are then calculated according to the proportional relationships between point pairs (including proportional relationships in quantity and distance). The other types of tower nodes (second, third, and auxiliary nodes) are scaled proportionally to obtain a new tower model. Finally, by comparing with the point cloud data, the model node positions are finely adjusted to optimize the results and improve model accuracy. This method accurately obtains the tower model corresponding to the tower point cloud data, achieving automatic modeling of tower point cloud data and solving the problems of high workload and poor feasibility associated with existing manual modeling.

[0122] In summary, the pole point cloud template matching method provided in this embodiment of the invention extracts pole point cloud data by scanning the target pole area with a 3D LiDAR scanner. Then, based on the point cloud shape, a corresponding template type is selected from a preset pole template database. Next, pole outline feature points are extracted from the pole point cloud data. This allows for automatic matching of these feature points with model feature points of the corresponding template type in the preset template database, resulting in the spatial position of the successfully matched model feature points. These model feature points then share the same spatial position as the pole outline feature points. By scaling the remaining model feature points proportionally, a complete pole model corresponding to the pole point cloud data can be obtained. This allows for matching the corresponding template type for each type of pole point cloud data, thus achieving automatic modeling. In conclusion, this method solves the problems of repetitive work and high time consumption in existing model building processes for similar pole point cloud data, enabling efficient construction of usable models corresponding to pole point cloud data and saving labor costs. In addition, it solves the problems of complex operation and too many parameters in the existing tower modeling process. By setting up the tower template database and matching the corresponding template type for targeted modeling, the manual operation cost can be greatly reduced, and the problem of large workload caused by various operation tasks and repetitive operations in the existing technology can be solved.

[0123] In addition, based on the same concept of the above method embodiments, the present invention also provides a template matching system for tower point clouds to implement the above method of the present invention. Since the principle and method of solving the problem in this system embodiment are similar, it has at least all the beneficial effects brought about by the technical solutions of the above embodiments, and will not be described in detail here.

[0124] See Figure 10 , Figure 10 This is a schematic diagram of the structure of a first type of pole point cloud template matching system provided in an embodiment of the present invention. Figure 10As shown, the template matching system for the tower point cloud includes:

[0125] The pole point cloud extraction module 110 is used to scan the target pole area with a 3D lidar and extract the pole point cloud data.

[0126] The template type selection module 120 is used to select the corresponding template type from the preset tower template database based on the point cloud shape of the tower point cloud data.

[0127] The contour feature extraction module 130 is used to extract the contour feature points of the tower from the tower point cloud data according to the template type.

[0128] The model feature matching module 140 is used to match tower outline feature points with model feature points of the corresponding template type in the preset tower template database to obtain the spatial position of the successfully matched model feature points.

[0129] The model feature scaling module 150 scales the remaining model feature points of the corresponding template type proportionally according to the spatial location of the successfully matched model feature points to obtain the tower model corresponding to the tower point cloud data.

[0130] As a preferred embodiment, the pole point cloud extraction module 110 is specifically used to scan the target power line corridor using a three-dimensional lidar to obtain point cloud data of the target power line corridor; to denoise and classify the point cloud data of the target power line corridor to obtain point cloud data of various power line types; and to extract pole point cloud data from the point cloud data of various power line types.

[0131] In a preferred embodiment, the template type selection module 120 is specifically used to select a pole image shape corresponding to the point cloud shape from a preset pole template database; determine the pole model corresponding to the pole image shape; search for the template type corresponding to the pole model from the preset pole template database; and extract the pole template corresponding to the template type.

[0132] In a preferred embodiment, the template type selection module 120 is further used to search for the pole template corresponding to the template type from the preset pole template database, and to take the inner and outer contour feature points of the pole template as the first type of model feature points; to obtain other types of pole nodes in the pole template based on the collinear relationship between each node in the pole template and the first type of model feature points; and to connect the first type of model feature points with other types of pole nodes to obtain the pole unit.

[0133] In a preferred embodiment, the contour feature extraction module 130 is specifically used to determine the tower template corresponding to the template type; extract the undulation structure of the first type of model feature points in the tower template, and the undulation structure of the tower contour in the tower point cloud data; wherein, the first type of model feature points are the inner and outer contour feature points of the tower template; perform position matching on the undulation structure of the first type of model feature points and the concave and convex points in the undulation structure of the tower contour; if the position matching of the concave and convex points in the undulation structure is successful, then extract the concave and convex points in the tower point cloud data as tower contour feature points.

[0134] In a preferred embodiment, the model feature matching module 140 is specifically used to select a tower template corresponding to a template type from a preset tower template database; perform feature matching between the tower outline feature points and the first type of model feature points in the tower template; when the tower outline feature points and the first type of model feature points are successfully matched, the spatial position of the tower outline feature points is taken as the spatial position of the matched first type of model feature points; and obtain the spatial positions of other types of tower nodes according to the distance ratio between other types of tower nodes in the tower template and the matched first type of model feature points.

[0135] In a preferred embodiment, the model feature matching module 140 is further configured to divide the first type of model feature points into coarse feature points and fine feature points according to the undulation structure of the first type of model feature points; perform feature matching between the tower outline feature points and the coarse feature points to obtain successfully matched coarse feature points; and obtain fine feature points that are successfully matched with the tower outline feature points according to the positional relationship between the successfully matched coarse feature points and the fine feature points, as well as the positional relationship between the fine feature points.

[0136] In a preferred embodiment, the model feature scaling module 150 is specifically used to translate the first type of model feature points to the spatial position of the corresponding tower outline feature points based on the feature matching results between the tower outline feature points and the first type of model feature points; and to scale the other types of tower nodes according to the point-to-point ratio between the other types of tower nodes and the first type of model feature points to obtain the tower model corresponding to the tower point cloud data.

[0137] In summary, the pole point cloud template matching system provided in this embodiment of the invention uses a three-dimensional LiDAR to scan the target pole area and extract pole point cloud data. Then, the pole type selection module 120 selects the corresponding template type from a preset pole template database based on the point cloud shape of the pole point cloud data. The contour feature extraction module 130 then extracts the pole contour feature points from the pole point cloud data according to the template type. In this way, the model feature matching module 140 can automatically match the pole contour feature points with the model feature points of the corresponding template type in the preset template database to obtain the spatial position of the successfully matched model feature points. Then, the model feature scaling module 150 scales the remaining model feature points proportionally based on the fact that the model feature points have the same spatial position as the pole contour feature points, thereby obtaining the complete pole model corresponding to the pole point cloud data. Thus, it is possible to match the corresponding template type for each type of pole point cloud data, thereby realizing automatic modeling.

[0138] In summary, the above method solves the problems of repetitive work and excessive time consumption in the existing model building process for similar pole point cloud data. It enables the efficient construction of usable models corresponding to pole point cloud data, thereby saving labor costs. Furthermore, it addresses the issues of complex operation and excessive parameter settings in the existing pole modeling process. By setting up a pole template database and matching corresponding template types for targeted modeling, the manual operation cost can be significantly reduced, resolving the problem of large workloads caused by various operational tasks and repetitive operations in existing technologies.

[0139] Additionally, see Figure 11 , Figure 11 This is a schematic diagram of the second type of template matching system for tower point clouds provided in an embodiment of the present invention. Figure 11 As shown, the template matching system for the tower point cloud includes:

[0140] The communication line 1002, communication module 1003, memory 1004, processor 1001, and a template matching program for pole point clouds stored in memory 1004 and executable on processor 1001, wherein when the template matching program for pole point clouds is executed by processor 1001, the steps of the template matching method for pole point clouds provided in any of the above embodiments are implemented.

[0141] In summary, the template matching scheme for pole point clouds described above, compared to existing technologies, solves the problems of repetitive work and excessive time consumption in the construction of identical pole point cloud models, enabling efficient construction of usable models and saving costs. Furthermore, it addresses the issues of complex operations and excessive parameter settings in pole modeling, allowing for a one-step model setup process that significantly reduces manual operation costs.

[0142] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0143] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0144] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0145] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0146] It should be noted that any reference signs placed between parentheses in the claims should not be construed as limiting the claims. The word "comprising" does not exclude the presence of components or steps not listed in the claims. The word "a" or "an" preceding a component does not exclude the presence of a plurality of such components. The invention can be implemented by means of hardware comprising several different components and by means of a suitably programmed computer. In a unit claim enumerating several means, several of these means may be embodied by the same item of hardware. The use of the words first, second, and third, etc., does not indicate any order. These words can be interpreted as names.

[0147] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0148] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method of template matching for a tower point cloud, the method comprising: include: The target tower area was scanned using a 3D LiDAR scanner to extract tower point cloud data. Based on the point cloud shape of the tower point cloud data, select the corresponding template type from the preset tower template database; Based on the template type, tower contour feature points are extracted from the tower point cloud data, specifically including: determining the tower template corresponding to the template type; extracting the undulation structure of the first type of model feature points in the tower template, and the undulation structure of the tower contour in the tower point cloud data; wherein, the first type of model feature points are the inner and outer contour feature points of the tower template; performing position matching between the concave and convex points in the undulation structure of the first type of model feature points and the concave and convex points in the undulation structure of the tower contour; if the position matching of the concave and convex points in the undulation structure is successful, then the concave and convex points in the tower point cloud data are extracted as the tower contour feature points; Matching the tower outline feature points with model feature points of the corresponding template type in the preset tower template database to obtain the spatial position of the successfully matched model feature points specifically includes: selecting a tower template corresponding to the template type from the preset tower template database; performing feature matching between the tower outline feature points and first-type model feature points in the tower template; when a tower outline feature point successfully matches a first-type model feature point, using the spatial position of the tower outline feature point as the spatial position of the matched first-type model feature point; and matching other types of tower nodes in the tower template with the matched first-type model feature points. The spatial positions of other types of tower nodes are obtained by calculating the distance ratio between feature points; based on the spatial positions of the successfully matched model feature points, the remaining model feature points of the corresponding template type are scaled proportionally to obtain the tower model corresponding to the tower point cloud data. Specifically, this includes: translating the first type of model feature points to the spatial positions of the corresponding tower contour feature points based on the feature matching results between the tower contour feature points and the first type of model feature points; and scaling the other types of tower nodes according to the point-to-point ratio between the other types of tower nodes and the first type of model feature points to obtain the tower model corresponding to the tower point cloud data.

2. The template matching method for tower point clouds according to claim 1, characterized in that, The step of using a 3D lidar to scan the target tower area and extract the tower point cloud data includes: The three-dimensional lidar is used to scan the target power line corridor to obtain point cloud data of the target power line corridor; The point cloud data of the target power line corridor is denoised and classified to obtain point cloud data of various power line types; The pole point cloud data is extracted from the point cloud data of the various power line types.

3. The method of template matching of tower point clouds of claim 1, wherein, The step of selecting a corresponding template type from a preset tower template database based on the point cloud shape of the tower point cloud data includes: Select a pole image shape that corresponds to the point cloud shape from the preset pole template database; Determine the tower model corresponding to the shape of the tower image; Search the preset pole template database for the template type corresponding to the pole model; Extract the tower template corresponding to the template type.

4. The method of template matching of tower point clouds of claim 1, wherein, Before the step of selecting a corresponding template type from a preset tower template database based on the point cloud shape of the tower point cloud data, the method further includes: Search the preset tower template database for the tower template type and use the inner and outer contour feature points of the tower template as the first type of model feature points; Based on the collinear relationship between each node in the tower template and the feature points of the first type of model, other types of tower nodes in the tower template are obtained; Connect the feature points of the first type of model with the other types of tower nodes to obtain tower units.

5. The method of template matching of tower point clouds of claim 4, wherein, The step of performing feature matching between the tower outline feature points and the first type of model feature points in the tower template includes: Based on the fluctuation structure of the feature points of the first type of model, the feature points of the first type of model are divided into coarse feature points and fine feature points; The tower outline feature points are matched with the coarse feature points to obtain successfully matched coarse feature points. Based on the positional relationship between the successfully matched coarse feature points and the fine feature points, as well as the positional relationship between the fine feature points, the fine feature points that successfully match the tower outline feature points are obtained.

Images