Wire generation method and device based on point cloud data, equipment and storage medium

Abstract

The embodiment of the invention provides a conductor generation method and device based on point cloud data, equipment and a storage medium, and relates to the field of three-dimensional modeling, and the method comprises the steps: obtaining the point cloud data, carrying out the classification processing of the point cloud data based on a weak supervised learning framework, obtaining the conductor point cloud data and the tower point cloud data, and obtaining the conductor point cloud data and the tower point cloud data; performing clipping processing on the conductor point cloud data according to the tower point cloud data to obtain clipped conductor point cloud data, performing classification processing on the clipped conductor point cloud data based on a DBSCAN clustering algorithm to obtain a training sample, and training a random forest model according to the training sample to obtain a conductor fitting model, the lead coordinate points are generated according to the tower point cloud data, the lead coordinate points are input into the lead fitting model, the lead list is output, the lead path in the point cloud data is automatically identified and extracted through the lead fitting model, the lead path extraction speed and accuracy are improved, and the lead path generation efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of three-dimensional modeling, and in particular to a method, device, equipment and storage medium for generating a wire based on point cloud data. Background Art

[0002] With the development of 3D laser scanning technology, point cloud data has become an important means of obtaining geospatial information. In urban planning, infrastructure construction and geographic information systems, point cloud data is increasingly used, such as in wire path generation.

[0003] Traditional wire path generation methods rely on manual operation, which is inefficient and easily affected by human factors, resulting in repeated wire path generation and difficult-to-control errors. In complex actual environments, traditional wire path generation methods require a lot of manpower and material resources, and cannot guarantee the accuracy of wire paths. Therefore, there is an urgent need for a method that can accurately generate wire paths based on point cloud data. Summary of the invention

[0004] In view of this, an object of the present invention is to overcome the deficiencies in the prior art and to provide a wire generation method, device, electronic device and readable storage medium based on point cloud data.

[0005] The present invention provides the following technical solutions:

[0006] In a first aspect, an embodiment of the present disclosure provides a method for generating a wire based on point cloud data, the method comprising:

[0007] Acquire point cloud data, and classify and process the point cloud data based on a weakly supervised learning framework to obtain conductor point cloud data and tower point cloud data;

[0008] Cutting the conductor point cloud data according to the tower point cloud data to obtain cut conductor point cloud data;

[0009] Classify the cut wire point cloud data based on the DBSCAN clustering algorithm to obtain training samples;

[0010] Training a random forest model according to the training samples to obtain a wire fitting model;

[0011] Generate conductor coordinate points according to the tower point cloud data, input the conductor coordinate points into the conductor fitting model, and output a conductor list.

[0012] According to a specific implementation method disclosed in the present application, the step of acquiring point cloud data, classifying and processing the point cloud data based on a weakly supervised learning framework, and obtaining conductor point cloud data and tower point cloud data includes:

[0013] Performing graffiti annotation processing on the conductors and towers in the point cloud data to obtain annotated point cloud data;

[0014] Generate pseudo labels according to the annotated point cloud data, and form a training set through the pseudo labels;

[0015] Training a neural network model according to the training set to obtain a preliminary classification model;

[0016] Iteratively optimizing the preliminary classification model to obtain a final classification model;

[0017] The point cloud data is input into the final classification model, and the conductor point cloud data and the tower point cloud data are output.

[0018] According to a specific implementation method disclosed in the present application, the step of clipping the conductor point cloud data according to the tower point cloud data to obtain clipped conductor point cloud data includes:

[0019] Determine the position of the tower and the vector between the towers according to the tower point cloud data;

[0020] Determine the size and direction of the clipping box according to the position of the pole tower and the vector between the pole towers;

[0021] The wire point cloud data is cropped according to the size and direction of the cropping box, and redundant data in the wire point cloud data is removed to obtain cropped wire point cloud data.

[0022] According to a specific implementation method disclosed in the present application, the step of classifying the cut wire point cloud data based on the DBSCAN clustering algorithm to obtain training samples includes:

[0023] Calculate the neighborhood of the midpoint of the clipping wire point cloud data;

[0024] Mark the core points in the clipping wire point cloud data according to the neighborhood;

[0025] Determine density connected points according to the core point and the neighborhood;

[0026] Marking noise points and boundary points in the cut wire point cloud data;

[0027] Cluster labels are added to the core points and the density-connected points, and the noise points are removed.

[0028] According to a specific implementation method disclosed in the present application, the step of classifying the cut wire point cloud data based on the DBSCAN clustering algorithm to obtain training samples also includes:

[0029] The input parameters of the DBSCAN clustering algorithm are las, eps and minSize, where las is the input clipping wire point cloud data, eps is the size of the calculation radius, and minSize is the minimum number of data points required around the core point.

[0030] According to a specific implementation method disclosed in the present application, the step of training a random forest model according to the training samples to obtain a wire fitting model includes:

[0031] Setting initialization parameters of the random forest model, and dividing the training samples into a training set and a validation set;

[0032] Training the random forest model according to the training set to obtain a preliminary wire fitting model;

[0033] The preliminary wire fitting model is iteratively optimized according to the verification set to obtain a wire fitting model.

[0034] According to a specific implementation method disclosed in the present application, the steps of generating conductor coordinate points according to the tower point cloud data, inputting the conductor coordinate points into the conductor fitting model, and outputting a conductor list include:

[0035] Determine the center point of the tower according to the tower point cloud data;

[0036] Determine the left boundary point and the right boundary point closest to the center point of the tower in the conductor point cloud data according to the center point of the tower;

[0037] Generate a wire coordinate point according to the distance between the left boundary point and the right boundary point;

[0038] According to the wire list formula P new,k =P left +NZ rf (P right -P left ) Get the coordinate points of the wire list, where P new,k is the coordinate point of the wire list, P right is the coordinate of the left boundary point, P left is the coordinate of the right boundary point, NZ rf Fit a model to the wire.

[0039] In a second aspect, an embodiment of the present disclosure provides a wire generation device based on point cloud data, the device comprising:

[0040] An acquisition module is used to acquire point cloud data, classify and process the point cloud data based on a weakly supervised learning framework, and obtain conductor point cloud data and tower point cloud data;

[0041] A cutting module, used for cutting the conductor point cloud data according to the tower point cloud data to obtain cut conductor point cloud data;

[0042] A classification module, used for classifying the cut wire point cloud data based on the DBSCAN clustering algorithm to obtain training samples;

[0043] A training module, used for training a random forest model according to the training samples to obtain a wire fitting model;

[0044] The generating module is used to generate the conductor coordinate points according to the tower point cloud data, input the conductor coordinate points into the conductor fitting model, and output the conductor list.

[0045] In a third aspect, an electronic device is provided in an embodiment of the present disclosure, the electronic device comprising a memory and a processor, the memory storing a computer program, and the processor implementing the steps of the wire generation method based on point cloud data described in any one of the first aspects when executing the computer program.

[0046] In a fourth aspect, a computer-readable storage medium is provided in an embodiment of the present disclosure, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the wire generation method based on point cloud data described in any one of the first aspects are implemented.

[0047] The above-mentioned method for generating a conductor based on point cloud data provided by the present application obtains point cloud data, classifies and processes the point cloud data based on a weakly supervised learning framework to obtain conductor point cloud data and pole tower point cloud data, crops the conductor point cloud data according to the pole tower point cloud data to obtain cropped conductor point cloud data, classifies and processes the cropped conductor point cloud data based on a DBSCAN clustering algorithm to obtain training samples, trains a random forest model based on the training samples to obtain a conductor fitting model, generates conductor coordinate points based on the pole tower point cloud data, inputs the conductor coordinate points into the conductor fitting model, outputs a conductor list, and automatically identifies and extracts conductor paths in point cloud data through the conductor fitting model, thereby improving the extraction speed and accuracy of the conductor path and the generation efficiency of the conductor path.

[0048] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and understandable, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments. It should be understood that the following drawings only illustrate certain embodiments of the present invention and should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work. In each of the drawings, similar components are numbered similarly.

[0050] Figure 1 A schematic diagram of a flow chart of a method for generating a wire based on point cloud data provided in an embodiment of the present application is shown;

[0051] Figure 2 A DBSCAN clustering wire result diagram provided in an embodiment of the present application is shown.

[0052] Figure 3 A schematic structural diagram of a wire generation device based on point cloud data provided in an embodiment of the present application is shown;

[0053] Figure 4 A schematic structural diagram of an electronic device provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0054] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be understood as limiting the present invention.

[0055] It should be noted that when an element is referred to as being "fixed to" another element, it may be directly on the other element or there may be a central element. When an element is considered to be "connected to" another element, it may be directly connected to the other element or there may be a central element at the same time. In contrast, when an element is referred to as being "directly on" another element, there is no intermediate element. The terms "vertical", "horizontal", "left", "right" and similar expressions used herein are for illustrative purposes only.

[0056] In the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, it can be the internal connection of two elements or the interaction relationship between two elements. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0057] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0058] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used herein in the specification of the template are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The term "and / or" used herein includes any and all combinations of one or more of the related listed items.

[0059] Example 1

[0060] like Figure 1 FIG. 1 is a flow chart of a method for generating a wire based on point cloud data provided in an embodiment of the present application. The method for generating a wire based on point cloud data provided in an embodiment of the present application includes the following steps:

[0061] Step S101, acquiring point cloud data, and classifying and processing the point cloud data based on a weakly supervised learning framework to obtain conductor point cloud data and tower point cloud data.

[0062] Specifically, point cloud data is obtained through laser radar, drone photogrammetry or other 3D scanning technology. During the acquisition of point cloud data, it is necessary to ensure that the data covers the entire area of ​​the wire to be generated. Point cloud data refers to a set of vectors in a 3D coordinate system. Each point in the point cloud data contains 3D coordinates.

[0063] The step of acquiring point cloud data, classifying and processing the point cloud data based on a weakly supervised learning framework, and obtaining conductor point cloud data and tower point cloud data includes:

[0064] Performing graffiti annotation processing on the conductors and towers in the point cloud data to obtain annotated point cloud data;

[0065] Generate pseudo labels according to the annotated point cloud data, and form a training set through the pseudo labels;

[0066] Training a neural network model according to the training set to obtain a preliminary classification model;

[0067] Iteratively optimizing the preliminary classification model to obtain a final classification model;

[0068] The point cloud data is input into the final classification model, and the conductor point cloud data and the tower point cloud data are output.

[0069] Specifically, graffiti is annotated on the wires and poles by using different colors or marks, such as red for the wires and blue for the poles. The annotated point cloud data is converted into digital form, such as identifying the annotated area by color segmentation or shape recognition. The points in the annotated point cloud data are classified by a machine learning classifier to generate pseudo labels. The pseudo labels are merged with the real annotated data to form a training set, and the training set is preprocessed to train the neural network model, iteratively optimize the model parameters to minimize the loss function, monitor the loss value during the training process and the accuracy on the validation set to evaluate the performance of the preliminary classification model. Adjust the structure, parameters or training strategy of the preliminary classification model, and retrain the model until the model performance is good to obtain the final classification model.

[0070] Step S102, cutting the conductor point cloud data according to the pole tower point cloud data to obtain cut conductor point cloud data.

[0071] The step of clipping the conductor point cloud data according to the tower point cloud data to obtain clipped conductor point cloud data comprises:

[0072] Determine the position of the tower and the vector between the towers according to the tower point cloud data;

[0073] Determine the size and direction of the clipping box according to the position of the pole tower and the vector between the pole towers;

[0074] The wire point cloud data is cropped according to the size and direction of the cropping box, and redundant data in the wire point cloud data is removed to obtain cropped wire point cloud data.

[0075] Specifically, the tower position is the precise three-dimensional coordinate of the tower in the tower point cloud data. The vector between the two towers can be obtained based on the three-dimensional coordinates of the two tower positions, and the vector can represent the direction of the wire. The size of the cropping box is set according to the tower position and the vector between the towers. The size can be adjusted according to the actual scene. The direction of the cropping box is parallel to the direction of the vector between the towers. The cropping box is constructed according to the tower position, the size and direction of the cropping box. Redundant data are points in the wire point cloud data that are not in the cropping box.

[0076] Step S103: classify the cut wire point cloud data based on the DBSCAN clustering algorithm to obtain training samples.

[0077] Specifically, there are usually different types and numbers of wires between towers. DBSCAN (Density-Based Spatial Clustering of Applications with Noise) is used for classification to provide data support for the subsequent wire fitting process. The DBSCAN algorithm has excellent robustness and automation features, can identify clusters of any shape, has strong noise processing capabilities, and can effectively deal with various proportions of noise and abnormal data points commonly found in point cloud data.

[0078] The step of classifying the cut wire point cloud data based on the DBSCAN clustering algorithm to obtain training samples includes:

[0079] Calculate the neighborhood of the midpoint of the clipping wire point cloud data;

[0080] Mark the core points in the clipping wire point cloud data according to the neighborhood;

[0081] Determine density connected points according to the core point and the neighborhood;

[0082] Marking noise points and boundary points in the cut wire point cloud data;

[0083] Cluster labels are added to the core points and the density-connected points, and the noise points are removed.

[0084] In specific implementation, for each point P in the clipped wire point cloud data, find all points whose distance to point P is less than ε, these points constitute the ε neighborhood of point P, and calculate the number of points in the ε neighborhood of each point P. If the number of points in the ε neighborhood of a point is greater than the preset minSize parameter, the point is marked as a core point. For each core point O, detect each point P in its ε neighborhood. If point P is also a core point, point P and point O are density-connected points. Noise points are points that are not included in the ε neighborhood of any core point, and boundary points are points that are density-connected to core points but are not core points.

[0085] The step of classifying the cut wire point cloud data based on the DBSCAN clustering algorithm to obtain training samples also includes:

[0086] The input parameters of the DBSCAN clustering algorithm are las, eps and minSize, where las is the input clipping wire point cloud data, eps is the size of the calculation radius, and minSize is the minimum number of data points required around the core point.

[0087] Specifically, see Figure 2 , Figure 2The DBSCAN clustering wire result diagram provided in the embodiment of the present application. The clustering result obtained by adjusting the eps parameter and the minSize parameter

[0088] Step S104: training a random forest model according to the training samples to obtain a wire fitting model.

[0089] The step of training the random forest model according to the training samples to obtain the wire fitting model comprises:

[0090] Setting initialization parameters of the random forest model, and dividing the training samples into a training set and a validation set;

[0091] Training the random forest model according to the training set to obtain a preliminary wire fitting model;

[0092] The preliminary wire fitting model is iteratively optimized according to the verification set to obtain a wire fitting model.

[0093] Understandably, the conductor does not always present a standard parabolic shape between towers, and traditional linear functions cannot accurately fit the conductor. The random forest algorithm can more accurately fit nonlinear conductors. By building multiple decision trees and taking the average of their results to reduce the risk of overfitting, while improving the generalization ability of the model, it can better capture the nonlinear characteristics of the conductor trajectory, improve the fitting accuracy and prediction ability of the model, and more accurately generate the actual path model of the conductor between towers.

[0094] Specifically, the wire fitting model consists of N decision trees, each tree T j Predict an elevation value z j Final prediction NZ rf is the wire fitting model of all tree fittings. The initialization parameters of the random forest include the number of trees, the maximum depth of the tree, and the complexity of the control tree. The performance of the preliminary wire fitting model is evaluated through the validation set to check the fitting degree and generalization ability of the model, and the model parameters are adjusted according to the evaluation results to achieve a better fitting effect and finally obtain the wire fitting model.

[0095] Step S105, generating conductor coordinate points according to the tower point cloud data, inputting the conductor coordinate points into the conductor fitting model, and outputting a conductor list.

[0096] The step of generating conductor coordinate points according to the tower point cloud data, inputting the conductor coordinate points into the conductor fitting model, and outputting a conductor list comprises:

[0097] Determine the center point of the tower according to the tower point cloud data;

[0098] Determine the left boundary point and the right boundary point closest to the center point of the tower in the conductor point cloud data according to the center point of the tower;

[0099] Generate a wire coordinate point according to the distance between the left boundary point and the right boundary point;

[0100] According to the wire list formula P new,k =P left +NZ rf (P right -P left ) Get the coordinate points of the wire list, where P new,k is the coordinate point of the wire list, P right is the coordinate of the left boundary point, P left is the coordinate of the right boundary point, NZ rf Fit a model to the wire.

[0101] Specifically, a distance formula, such as Euclidean distance, is used to determine the left boundary point and the right boundary point closest to the center point of the pole tower, and a series of conductor coordinate points are generated according to the left boundary point and the right boundary point, and the conductor coordinate points are evenly spaced between the left boundary point and the right boundary point. The conductor list coordinate points are sorted into a preset format and returned from the back-end interface.

[0102] The wire generation method based on point cloud data provided in the present embodiment obtains point cloud data, classifies and processes the point cloud data based on a weakly supervised learning framework to obtain wire point cloud data and pole tower point cloud data, crops the wire point cloud data according to the pole tower point cloud data to obtain cropped wire point cloud data, classifies and processes the cropped wire point cloud data based on a DBSCAN clustering algorithm to obtain training samples, trains a random forest model based on the training samples to obtain a wire fitting model, generates wire coordinate points based on the pole tower point cloud data, inputs the wire coordinate points into the wire fitting model, outputs a wire list, and automatically identifies and extracts wire paths in point cloud data through the wire fitting model, thereby improving the extraction speed and accuracy of wire paths and the generation efficiency of wire paths.

[0103] Example 2

[0104] like Figure 3 , which is a schematic diagram of a structure of a wire generating device 300 based on point cloud data provided in an embodiment of the present application, wherein the device comprises:

[0105] An acquisition module 301 is used to acquire point cloud data, classify and process the point cloud data based on a weakly supervised learning framework, and obtain conductor point cloud data and tower point cloud data;

[0106] A cutting module 302 is used to cut the conductor point cloud data according to the tower point cloud data to obtain cut conductor point cloud data;

[0107] A classification module 303 is used to classify the cut wire point cloud data based on the DBSCAN clustering algorithm to obtain training samples;

[0108] A training module 304 is used to train a random forest model according to the training samples to obtain a wire fitting model;

[0109] The generating module 305 is used to generate conductor coordinate points according to the tower point cloud data, input the conductor coordinate points into the conductor fitting model, and output a conductor list.

[0110] The point cloud data-based wire generation device 300 provided in this embodiment can implement the point cloud data-based wire generation method shown in Example 1, and will not be described again here to avoid repetition.

[0111] The conductor generation device based on point cloud data provided in the embodiment of the present application obtains point cloud data, classifies and processes the point cloud data based on a weakly supervised learning framework to obtain conductor point cloud data and pole tower point cloud data, crops the conductor point cloud data according to the pole tower point cloud data to obtain cropped conductor point cloud data, classifies and processes the cropped conductor point cloud data based on a DBSCAN clustering algorithm to obtain training samples, trains a random forest model based on the training samples to obtain a conductor fitting model, generates conductor coordinate points based on the pole tower point cloud data, inputs the conductor coordinate points into the conductor fitting model, outputs a conductor list, and automatically identifies and extracts conductor paths in point cloud data through the conductor fitting model, thereby improving the extraction speed and accuracy of the conductor path and the generation efficiency of the conductor path.

[0112] Example 3

[0113] In addition, an embodiment of the present disclosure provides an electronic device 400, including a memory and a processor, wherein the memory stores a computer program.

[0114] For details, see Figure 4 The electronic device 400 includes: a receiver 401, a bus interface and a processor 402.

[0115] In the embodiment of the present invention, the electronic device 400 further includes a memory 403. Figure 4In the embodiment, the bus architecture may include any number of interconnected buses and bridges, specifically linking various circuits of one or more processors represented by processor 402 and memory represented by memory 403. The bus architecture may also link various other circuits such as peripherals, voltage regulators, and power management circuits, which are well known in the art and are therefore not further described herein. The bus interface provides an interface. The receiver 401 may be a plurality of components, namely, a transmitter and a receiver, providing a unit for communicating with various other devices on a transmission medium. The processor 402 is responsible for managing the bus architecture and general processing, and the memory 403 may store data used by the processor 402 when performing operations.

[0116] The electronic device 400 provided in the embodiment of the present invention can implement the wire generation method based on point cloud data shown in Example 1, which will not be described again to avoid repetition.

[0117] In several embodiments provided in the present application, it should be understood that the disclosed devices and methods can also be implemented in other ways. The device embodiments described above are merely schematic. For example, the flowcharts and structure diagrams in the accompanying drawings show the possible architecture, functions and operations of the devices, methods and computer program products according to multiple embodiments of the present invention. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of a code, and the module, a program segment or a part of a code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in an alternative implementation, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the structure diagram and / or the flow chart, and the combination of boxes in the structure diagram and / or the flow chart, can be implemented with a dedicated hardware-based system that performs a specified function or action, or can be implemented with a combination of dedicated hardware and computer instructions.

[0118] In addition, the functional modules or units in the various embodiments of the present invention may be integrated together to form an independent part, or each module may exist independently, or two or more modules may be integrated to form an independent part.

[0119] If the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a smart phone, a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0120] The above description is only a specific implementation mode of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A method for generating a wire based on point cloud data, characterized in that: The method comprises: Acquire point cloud data, and classify and process the point cloud data based on a weakly supervised learning framework to obtain conductor point cloud data and tower point cloud data; Cutting the conductor point cloud data according to the tower point cloud data to obtain cut conductor point cloud data; Classify the cut wire point cloud data based on the DBSCAN clustering algorithm to obtain training samples; Training a random forest model according to the training samples to obtain a wire fitting model; Generate conductor coordinate points according to the tower point cloud data, input the conductor coordinate points into the conductor fitting model, and output a conductor list.

2. The method for generating a wire based on point cloud data according to claim 1, characterized in that: The step of acquiring point cloud data, classifying and processing the point cloud data based on a weakly supervised learning framework, and obtaining conductor point cloud data and tower point cloud data includes: Performing graffiti annotation processing on the conductors and towers in the point cloud data to obtain annotated point cloud data; Generate pseudo labels according to the annotated point cloud data, and form a training set through the pseudo labels; Training a neural network model according to the training set to obtain a preliminary classification model; Iteratively optimizing the preliminary classification model to obtain a final classification model; The point cloud data is input into the final classification model, and the conductor point cloud data and the tower point cloud data are output.

3. The method for generating a wire based on point cloud data according to claim 1, characterized in that: The step of clipping the conductor point cloud data according to the tower point cloud data to obtain clipped conductor point cloud data comprises: Determine the position of the tower and the vector between the towers according to the tower point cloud data; Determine the size and direction of the clipping box according to the position of the pole tower and the vector between the pole towers; The wire point cloud data is cropped according to the size and direction of the cropping box, and redundant data in the wire point cloud data is removed to obtain cropped wire point cloud data.

4. The method for generating a wire based on point cloud data according to claim 1, characterized in that: The step of classifying the cut wire point cloud data based on the DBSCAN clustering algorithm to obtain training samples includes: Calculate the neighborhood of the midpoint of the clipping wire point cloud data; Mark the core points in the clipping wire point cloud data according to the neighborhood; Determine density connected points according to the core point and the neighborhood; Marking noise points and boundary points in the cut wire point cloud data; Cluster labels are added to the core points and the density-connected points, and the noise points are removed.

5. The method for generating a wire based on point cloud data according to claim 1, characterized in that: The step of classifying the cut wire point cloud data based on the DBSCAN clustering algorithm to obtain training samples also includes: The input parameters of the DBSCAN clustering algorithm are las, eps and minSize, where las is the input clipping wire point cloud data, eps is the size of the calculation radius, and minSize is the minimum number of data points required around the core point.

6. The method for generating a wire based on point cloud data according to claim 1, characterized in that: The step of training the random forest model according to the training samples to obtain the wire fitting model comprises: Setting initialization parameters of the random forest model, and dividing the training samples into a training set and a validation set; Training the random forest model according to the training set to obtain a preliminary wire fitting model; The preliminary wire fitting model is iteratively optimized according to the verification set to obtain a wire fitting model.

7. The method for generating a wire based on point cloud data according to claim 1, characterized in that: The step of generating conductor coordinate points according to the tower point cloud data, inputting the conductor coordinate points into the conductor fitting model, and outputting a conductor list comprises: Determine the center point of the tower according to the tower point cloud data; Determine the left boundary point and the right boundary point closest to the center point of the tower in the conductor point cloud data according to the center point of the tower; Generate a wire coordinate point according to the distance between the left boundary point and the right boundary point; According to the wire list formula P new,k =P left +NZ rf (P right -P left ) Get the coordinate points of the wire list, where P new,k is the coordinate point of the wire list, P right is the coordinate of the left boundary point, P left is the coordinate of the right boundary point, BZ rf Fit a model to the wire.

8. A wire generation device based on point cloud data, characterized in that: The device comprises: An acquisition module is used to acquire point cloud data, classify and process the point cloud data based on a weakly supervised learning framework, and obtain conductor point cloud data and tower point cloud data; A cutting module, used for cutting the conductor point cloud data according to the tower point cloud data to obtain cut conductor point cloud data; A classification module, used for classifying the cut wire point cloud data based on the DBSCAN clustering algorithm to obtain training samples; A training module, used for training a random forest model according to the training samples to obtain a wire fitting model; The generating module is used to generate the conductor coordinate points according to the tower point cloud data, input the conductor coordinate points into the conductor fitting model, and output the conductor list.

9. An electronic device, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the method for generating a wire based on point cloud data according to any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method for generating a wire based on point cloud data according to any one of claims 1 to 7 are implemented.

Images