Road slope three-dimensional visual modeling method, device, medium and equipment

Abstract

The invention discloses a road slope three-dimensional visualization modeling method and device, a medium and equipment, and relates to the technical field of three-dimensional visualization. Comprising the following steps: collecting point cloud data of a road slope, and obtaining a feature vector of each point based on the point cloud data; dividing each point into a plurality of different classes based on the similarity of the feature vectors, wherein each class comprises a plurality of sub-clusters dispersed in geographic space positions; obtaining a confusion feature value used for representing the confusion degree of the point cloud data in each sub-cluster and the number of sub-clusters belonging to other categories in the neighborhood of each sub-cluster; setting the number of triangular meshes in direct proportion to the disordered feature value and the number of the sub-clusters belonging to other categories for each sub-cluster; and on the basis of the set number of the triangular meshes, using a triangulation method to connect the midpoints of the sub-clusters into non-overlapping triangular meshes, and obtaining a road slope three-dimensional model.

Description

Technical Field

[0001] The invention relates to the field of three-dimensional visualization technology, and in particular to a three-dimensional visualization modeling method, device, medium and equipment for a highway slope. Background Art

[0002] Highway slope 3D visualization modeling is a process of converting the actual scene of highway slope into a 3D model and realizing visualization through computer technology, which provides a powerful tool for the design, analysis, monitoring and maintenance of highway slope. When the highway slope 3D visualization modeling is carried out, the point cloud data of the highway slope surface is first collected. Then, the points in the point cloud data are connected into a grid surface by triangulation of different sizes to draw the surface of the highway slope 3D model. Finally, the real scene photos are pasted into the corresponding grid to obtain a 3D model of the highway slope with texture.

[0003] In the prior art, Chinese patent CN112734924B discloses a method for visual modeling of a real-scene 3D model of a roadbed slope, which includes: importing real-scene photos and real-scene videos with satellite positioning information; aligning real-scene photos; generating dense point clouds based on the aligned real-scene photos; generating grids based on the dense point clouds; generating textures based on the grids; exporting a standard file of the 3D model of the roadbed slope; and lightly loading the standard file of the 3D model of the roadbed slope. When generating the grid, the method artificially sets the number of grid faces to 30,000.

[0004] However, the above method manually sets the number of grids based on experience, which may cause the set number of grids to not match the point cloud data, making it difficult to apply to different types of highway slope models. Summary of the invention

[0005] Based on this, in order to solve the technical problems in the prior art, the present invention provides a three-dimensional visualization modeling method, device, medium and equipment for highway slopes.

[0006] The present invention provides a three-dimensional visualization modeling method for a highway slope, comprising:

[0007] Collecting point cloud data of the highway slope, wherein the point cloud data includes the three-dimensional coordinates of each point and the signal reflection characteristic value of each point;

[0008] Extracting the feature vector of each point based on the three-dimensional coordinates and signal reflection feature values ​​of each point, and classifying each point into several different classes according to the similarity of the feature vectors, each class includes several sub-clusters dispersed in geographic space;

[0009] Obtaining a chaotic eigenvalue for characterizing the degree of chaos of the point cloud data in each subcluster and the number of subclusters belonging to other categories in the neighborhood of each subcluster; setting the number of triangular meshes for each subcluster that is proportional to the chaotic eigenvalue and the number of subclusters belonging to other categories;

[0010] Combined with the set number of triangular meshes, the midpoints of each sub-cluster are connected into non-overlapping triangular meshes using the triangulation method to obtain the three-dimensional model of the highway slope.

[0011] Furthermore, the step of setting the number of triangular meshes for each subcluster that is proportional to the chaotic eigenvalue and the number of subclusters belonging to other categories specifically includes:

[0012] Get the number of base grids N T :

[0013] N T =n x ×n y ×n z

[0014]

[0015] Among them, n x 、n y and n z are the maximum number of grids in the horizontal, vertical and height dimensions respectively; Ranye x 、Range y and Range z are the spatial spans of the horizontal, vertical and height dimensions respectively; max x and min x Respectively represent the maximum and minimum coordinate positions of the horizontal dimension, max y and min y Respectively represent the maximum and minimum coordinate positions of the longitudinal dimension; max z and min z Respectively represent the maximum and minimum coordinate positions of the height dimension; F is the minimum number of point cloud data contained in each grid; Indicates rounding up;

[0016] Get the center of each subcluster in each class, and get the number of subclusters of other classes in the ε neighborhood of each subcluster center represents the number of subclusters that do not belong to the i-th category in the ε neighborhood of the j-th subcluster in the i-th category, ε = max(n x ,n y ,n z );

[0017] The variance of the eigenvector in each sub-cluster is used as the chaotic eigenvalue of the degree of chaos of the point cloud data in the corresponding sub-cluster

[0018] based on and The product of is proportional to the number of base grids N T , get the number of triangle meshes corresponding to each subcluster:

[0019]

[0020] in, is the number of triangle meshes corresponding to the jth subcluster in the i-th class; α is the adjustment coefficient.

[0021] Furthermore, the points are divided into several different classes according to the similarity of the feature vectors, specifically including:

[0022] The feature vectors are clustered using density-based spatial clustering algorithms and k-means clustering algorithms, and corresponding category labels are added to each point according to the clustering results:

[0023]

[0024] Among them, p i is the i-th point in the highway slope point cloud data; It is a density-based spatial clustering algorithm for p i The first category label added, k is the total number of class labels; is the k-means clustering algorithm for p i Added second category label,

[0025] Points with the same first and second category labels are classified into one category, so as to reallocate category labels to each point, and obtain several different categories based on the reallocated category labels.

[0026] Furthermore, extracting the feature vector of each point based on the three-dimensional coordinates of each point and the signal reflection feature value specifically includes:

[0027] For any point p in the point cloud dataset i (x i ,y i ,z i ), extract its signal reflection characteristic value θ i and the point set N(p i ):

[0028]

[0029] in, It's point p i d points in the neighborhood δ;

[0030] Use local least squares fitting to fit each point p i and its neighborhood N(p i ) and take the quadratic surface as point p i The corresponding quadratic curvature;

[0031] Point p i Neighborhood N(p i ) to obtain the corresponding quadratic surface curvature N(k i ):

[0032]

[0033] in, They are point p i The curvature of the quadratic surface corresponding to the d points in the neighborhood δ;

[0034] θ i , k i and N(k i ) to obtain any point p i The eigenvector P i ={θ i ,k i ,N(k i )}.

[0035] The present invention provides a three-dimensional visual modeling device for a highway slope, comprising:

[0036] A data acquisition module, used to collect point cloud data of the highway slope, wherein the point cloud data includes the three-dimensional coordinates of each point and the signal reflection characteristic value of each point;

[0037] A clustering module is used to extract the feature vector of each point based on the three-dimensional coordinates and signal reflection feature values ​​of each point, and to classify each point into several different classes according to the similarity of the feature vectors. Each class includes several sub-clusters dispersed in geographical space.

[0038] A triangular mesh number acquisition module is used to obtain a chaotic eigenvalue used to characterize the degree of chaos of point cloud data in each subcluster and the number of subclusters belonging to other categories in the neighborhood of each subcluster; and to set the number of triangular meshes for each subcluster that is proportional to the chaotic eigenvalue and the number of subclusters belonging to other categories;

[0039] The modeling module is used to connect the midpoints of each sub-cluster into non-overlapping triangular meshes using a triangulation method in combination with the set number of triangular meshes to obtain a three-dimensional model of the highway slope.

[0040] The present invention provides a computer-readable storage medium, wherein the storage medium stores a computer program, and when the computer program is executed by a processor, the above-mentioned three-dimensional visualization modeling method of highway slope is implemented.

[0041] The present invention provides a computer device, comprising a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the above-mentioned three-dimensional visualization modeling method of highway slope when executing the program.

[0042] At least one of the above technical solutions adopted by the present invention can achieve the following beneficial effects:

[0043] In the three-dimensional visualization modeling method for highway slopes provided by the present invention, the complexity of the point cloud data inside and outside each subcluster is obtained through the chaotic eigenvalue of each subcluster and the number of subclusters belonging to other categories in the neighborhood of each subcluster, and the number of triangular meshes proportional to the chaotic eigenvalue and the number of subclusters belonging to other categories is set for each subcluster, so that the number of triangular meshes is adapted to the complexity of the point cloud data, avoiding the problem of excessive or insufficient meshes caused by the mismatch between the number of meshes and the complexity of the point cloud data due to artificial settings, thereby being able to adapt to point cloud data corresponding to different types of highway slopes. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0045] Figure 1 A schematic diagram of a process flow of a three-dimensional visual modeling method for a highway slope provided by the present invention;

[0046] Figure 2 A schematic diagram of a highway slope provided by the present invention;

[0047] Figure 3 A schematic diagram of a process for obtaining the number of triangular meshes of a sub-cluster provided by the present invention;

[0048] Figure 4 A schematic diagram of a computer device for realizing a three-dimensional visual modeling method of a highway slope provided by the present invention.

[0049] Attached tags:

[0050] (1) Retaining wall; (2) Soil. DETAILED DESCRIPTION

[0051] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in combination with the specific embodiments of the present invention and the corresponding drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments in the specification, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0052] In the process of 3D visualization modeling of highway slopes, the number of triangular meshes is directly related to the 3D model of highway slopes: a small number of mesh faces may lead to an oversimplified model that cannot accurately reflect the true shape and detailed features of the highway slopes; while too many mesh faces may lead to an overly complex and redundant model, increasing the difficulty of calculation and processing. In addition, the number of mesh faces directly affects the rendering efficiency of the 3D model: if the number of mesh faces is too large, the rendering time will increase significantly, which may make real-time rendering difficult, which not only affects the user experience, but also may limit the use of the model in real-time applications. Therefore, there is an urgent need for a 3D visualization modeling method for highway slopes that can adaptively set the appropriate number of meshes according to the characteristics of point cloud data.

[0053] Based on this, the present invention provides a three-dimensional visualization modeling method for highway slopes. The method uses the curvature around each point to represent the geographic spatial feature corresponding to the point, and uses the signal reflection characteristic value as the geological feature corresponding to the point, thereby obtaining the characteristic vector of each point; based on the characteristic vector, points with similar geographic spatial features and geological features are clustered into one category, and then the degree of confusion of the characteristic vector and the number of point clouds in each category are used to set the corresponding triangular mesh data: the lower the degree of confusion of the characteristic vector and the more triangular mesh data, the stronger the homogeneity of the point cloud data in this category, and the lower the corresponding number of meshes. The method can realize the adaptive setting of the number of meshes based on the characteristics of the point cloud data.

[0054] Example 1

[0055] Figure 1 The process of the three-dimensional visualization modeling method of highway slope in this embodiment is shown. Figure 1 The method is described in detail and specifically comprises the following steps:

[0056] S1: Collecting point cloud data of the road slope, wherein the point cloud data includes the three-dimensional coordinates of each point and the signal reflection characteristic value of each point.

[0057] The actual highway slope is used as the aerial survey area, and a drone equipped with a camera or radar is used to collect data in the aerial survey area according to a preset route. The data obtained from the aerial survey is sorted to obtain point cloud data of the highway slope surface for three-dimensional reconstruction. The point cloud data depicts the highway slope through a series of discrete points, which represent the measurement points on the highway slope surface. Each point cloud data includes:

[0058] p i ={(x i ,y i ,z i ), θ i}

[0059] Among them, (x i ,y i ,z i ) is point p i The coordinate position in the horizontal, vertical and height directions in three-dimensional space; θ i For point p i The signal reflection characteristic values ​​include reflection intensity and echo number. Reflection intensity refers to the intensity of the laser beam emitted by the lidar reflected from the surface of an object. Different surface materials and structures of objects have different reflection intensities for the laser. Therefore, the reflection intensity information can be used to distinguish different objects or materials. The laser beam emitted by the lidar may produce multiple reflections on the surface of an object to form multiple echo signals. The echo number information can reflect the degree of chaos and depth information of the object's surface.

[0060] S2: Extract the feature vector of each point based on the three-dimensional coordinates and signal reflection eigenvalues ​​of each point, and divide each point into several different classes according to the similarity of the feature vectors. Each class includes several subclusters dispersed in geographic space.

[0061] S201: Obtain the feature vector of each point.

[0062] For any point p in the point cloud dataset i (x i ,y i ,z i ), extract the point set in its corresponding neighborhood in, It's point p i There are d points in the neighborhood δ.

[0063] Curvature is the degree of curvature of the surface around a point. It is used to describe the undulation of the terrain or the change of the surface. By calculating the change of curvature in the neighborhood of each point, the terrain characteristics of the area where the point is located can be identified. Use local least squares to fit each point p i and its neighborhood N(p i ) and obtain the corresponding quadratic surface of point p i The Gaussian curvature k of the corresponding quadratic surface i , and its neighborhood N(p i ) corresponding to the Gaussian curvature of the quadratic surface They are point p i The curvature of the quadratic surface corresponding to the d points in the neighborhood δ.

[0064] k i and represents point p i The corresponding geographic spatial features are expressed as θ i As point p i Corresponding geological features, θ i , k i and N(k i ) to obtain any point p i The eigenvector P i ={θ i ,k i ,N(k i )}, the curvature change and signal reflection eigenvalue of each point are combined into a comprehensive feature vector, through which the geographic information and surface features of each point in the point cloud can be more comprehensively represented.

[0065] S202: clustering points with similar feature vectors into one class to obtain a number of different classes, each class including a number of sub-clusters dispersed in geographic space.

[0066] The feature vectors are clustered using density-based spatial clustering algorithms and k-means clustering algorithms, and corresponding category labels are added to each point according to the clustering results:

[0067]

[0068] Among them, p i is the i-th point in the highway slope point cloud data; It is a density-based spatial clustering algorithm for p i The first category label added, k is the total number of class labels; is the k-means clustering algorithm for p i Added second category label,

[0069] To facilitate understanding of sub-clusters, Figure 2Taking the highway slope shown as an example, the highway slope includes a retaining wall (1) for preventing the soil (2) from sliding and the soil (2) located between the retaining wall (1). Based on the similarity of the feature vectors, the point cloud data corresponding to the highway slope at least includes the corresponding retaining wall category and the corresponding soil category. In the point cloud data of the highway slope, the retaining wall may be composed of multiple areas, each of which may have a certain degree of dispersion in space. Therefore, the point cloud data in the retaining wall category is not a unified whole, but is composed of multiple subclusters; each subcluster represents a retaining wall area, which may be located in different geographical locations and even present different structural forms. Similarly, the soil may also be composed of multiple scattered point cloud areas, each of which represents a soil area, and these areas locally present different forms of soil characteristics.

[0070] The core idea of ​​DBSCAN (Density-Based Spatial Clustering of Applications with Noise) is to define clusters by finding high-density areas in the data set, and automatically cluster and identify noise according to the density relationship between data points. K-means clustering is a partition-based clustering algorithm that iteratively assigns data points to one of k clusters and optimizes the clustering results by minimizing the distance from each data point to the cluster center. The two clustering algorithms assign a category label to each point cloud data, indicating which cluster the point belongs to. Two clustering algorithms are used to analyze the same data set from multiple angles. If both clustering algorithms classify some data points into the same class, then these points have strong similarities. Points with the same first and second category labels are classified into one category to reassign category labels to each point, and several different classes are obtained based on the reassigned category labels. Table 1 shows examples of the first category labels, second category labels, and reassigned category labels for some point cloud data.

[0071] Table 1 The first category labels, second category labels and reassigned category labels of point cloud data

[0072] Point cloud data number First Category Label Second category label Reassigned class labels 1 2 3 1 .... .... .... .... 500 2 3 1 501 3 4 2 .... .... .... .... 1000 3 4 2 1001 4 k k .... .... .... .... Num 4 k k

[0073] By redistributing the point cloud data, new categories can be generated based on label consistency. This process helps to enhance the clustering of similar point cloud data, making the data points in the same category closer and more consistent. If the point cloud data is clustered by two different algorithms, and some data points are assigned to the same cluster in both algorithms, then it can be considered that there is a strong similarity between these data points. This indicates that they may correspond to the same geomorphic feature or the same type of spatial structure. Therefore, these data points are considered to belong to the same type of geomorphic and can provide consistent geographic or spatial information. Conversely, if some data points are assigned to different clusters in different algorithms, this indicates that the similarity between these points is weak, which may mean that they belong to different geomorphic features or spatial structures. In this case, the classification results of these data points may be inconsistent, resulting in them not having strong similarity, so they can be considered to correspond to different geomorphic features.

[0074] Step S2 obtains a feature vector that can more comprehensively represent the geographic information and surface features of each point in the point cloud through the curvature change and signal reflection eigenvalue in the neighborhood of each point; based on the feature vector, different clustering methods are used for classification, and the category labels are reallocated according to the classification results, so that the probability that the same type of point cloud data corresponds to the same highway slope landform feature is greater.

[0075] S3: Obtain the chaotic eigenvalue used to characterize the degree of chaos of the point cloud data in each subcluster and the number of subclusters belonging to other categories in the neighborhood of each subcluster; set the number of triangular meshes for each subcluster that is proportional to the chaotic eigenvalue and the number of subclusters belonging to other categories. Figure 3 As shown, specifically including:

[0076] S301: Setting the number of reference grids N T .

[0077] Set the minimum number of point cloud data F contained in each grid, that is, each grid is expected to contain at least F points. Based on the minimum number of point cloud data F, the maximum number of grids corresponding to the highway slope point cloud data in three dimensions is obtained:

[0078] n x =Range x / F=(max x -min x ) / F

[0079] n y =Range y / F=(max y -min y ) / F

[0080] n z =Rangez / F=(max z -min z ) / F

[0081] Among them, n x 、n y and n z The maximum number of grids in the horizontal, vertical and height dimensions respectively; Range x 、Range y and Range z are the spatial spans of the horizontal, vertical and height dimensions respectively; max x and min x Respectively represent the maximum and minimum coordinate positions in the horizontal dimension, max y and min y Respectively represent the maximum and minimum coordinate positions in the vertical dimension; max z and min z Represent the maximum and minimum coordinate positions in the height dimension respectively. x 、n y and n z Multiply to get the number of base grids N T =n x ×n y ×n z .

[0082] S302: Obtaining a chaos feature value for characterizing the degree of chaos of point cloud data in each sub-cluster.

[0083] The variance of the eigenvector in each sub-cluster is used as the chaotic eigenvalue of the degree of chaos of the point cloud data in the corresponding sub-cluster The chaotic eigenvalue reflects the degree of chaos of the point cloud data from within the subcluster. The larger the variance of the eigenvector, the higher the degree to which the eigenvector in the subcluster deviates from the overall average level, and the greater the degree of chaos of the point cloud data in the subcluster.

[0084] S303: Obtain the number of subclusters belonging to other categories in the neighborhood of each subcluster.

[0085] In step 301, the maximum number of grids corresponding to the three dimensions (n x ,n y ,n z ), and based on the number of grids, set the search neighborhood ε=max(n x ,n y ,n z ). Then obtain the center of each subcluster in each class, and count the number of subclusters of other classes within the ε neighborhood of each subcluster center It represents the number of subclusters in the ε neighborhood of the jth subcluster in the i-th category that do not belong to the i-th category.

[0086] S304: Obtaining the number of triangle meshes corresponding to each sub-cluster

[0087]

[0088] in, is the number of triangle meshes corresponding to the jth subcluster in the i-th class; α is the adjustment coefficient.

[0089] S4: Based on the set number of triangular meshes, the midpoints of each sub-cluster are connected into non-overlapping triangular meshes using a triangulation method to obtain a three-dimensional model of the highway slope.

[0090] Each subcluster has its corresponding point cloud data and number of triangular meshes. Delaunay triangulation is performed on the points in each class, and each subcluster's slope area will have a corresponding triangular mesh model. Finally, these models are merged to obtain a 3D model of the entire highway slope.

[0091] based on Figure 1 The three-dimensional visualization modeling method of highway slope shown in the figure reflects the distribution differences of point clouds within each category through chaotic eigenvalues, providing a basis for grid division, so that the number of triangular meshes is adapted to the complexity of the data within the category; by dynamically setting the number of triangular meshes according to the characteristic fluctuations of each category, it can adapt to different point cloud data types, avoid excessive or insufficient meshing problems caused by the mismatch between mesh density and data complexity, and can accurately control the complexity of the mesh, thereby maintaining sufficient details and improving computational efficiency.

[0092] The above is a three-dimensional visualization modeling method for highway slopes provided by one or more embodiments of the present invention. Based on the same idea, the present invention also provides a corresponding three-dimensional visualization modeling device for highway slopes, including:

[0093] The data acquisition module is used to collect point cloud data of the highway slope, wherein the point cloud data includes the three-dimensional coordinates of each point and the signal reflection characteristic value of each point.

[0094] The clustering module is used to extract the feature vector of each point based on the three-dimensional coordinates and signal reflection eigenvalues ​​of each point, and classify each point into several different classes according to the similarity of the feature vectors. Each class includes several sub-clusters dispersed in geographic space.

[0095] The triangular mesh number acquisition module is used to obtain the chaotic eigenvalue used to characterize the degree of chaos of the point cloud data in each subcluster and the number of subclusters belonging to other categories in the neighborhood of each subcluster; and set the number of triangular meshes for each subcluster that is proportional to the chaotic eigenvalue and the number of subclusters belonging to other categories.

[0096] The modeling module is used to connect the midpoints of each sub-cluster into non-overlapping triangular meshes using a triangulation method in combination with the set number of triangular meshes to obtain a three-dimensional model of the highway slope.

[0097] For the specific definition of the highway slope three-dimensional visualization modeling device, please refer to the definition of the highway slope three-dimensional visualization modeling method above, which will not be repeated here. Each module in the above-mentioned highway slope three-dimensional visualization modeling device can be implemented in whole or in part by software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0098] The present invention also provides a computer-readable storage medium, which stores a computer program, which can be used to execute the above Figure 1 The proposed method is a three-dimensional visualization modeling method for highway slopes.

[0099] The present invention also provides a computer device structure. At the hardware level, the computer device includes a processor, an internal bus, a network interface, a memory, and a non-volatile memory. Of course, it may also include other hardware required for the business. The processor reads the corresponding computer program from the non-volatile memory into the memory and then runs it to achieve the above Figure 1 The proposed method is a three-dimensional visualization modeling method for highway slopes.

[0100] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided by the present invention can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory or optical memory, etc. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0101] The technical features of the above embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of the present invention.

Claims

1. A three-dimensional visual modeling method for highway slopes, characterized in that: include: Collecting point cloud data of the highway slope, wherein the point cloud data includes the three-dimensional coordinates of each point and the signal reflection characteristic value of each point; Extracting the feature vector of each point based on the three-dimensional coordinates and signal reflection feature values ​​of each point, and classifying each point into several different classes according to the similarity of the feature vectors, each class includes several sub-clusters dispersed in geographic space; Obtaining the chaotic eigenvalues ​​used to characterize the degree of chaos of the point cloud data in each subcluster and the number of subclusters belonging to other categories in the neighborhood of each subcluster; For each subcluster, the number of triangle meshes is set proportional to the chaotic eigenvalue and the number of subclusters belonging to other categories; Combined with the set number of triangular meshes, the midpoints of each sub-cluster are connected into non-overlapping triangular meshes using the triangulation method to obtain the three-dimensional model of the highway slope.

2. The three-dimensional visual modeling method of highway slope according to claim 1, characterized in that: The step of setting the number of triangular meshes for each subcluster that is proportional to the chaotic eigenvalue and the number of subclusters belonging to other categories specifically includes: Get the number of base grids N T : N T =n x ×n y ×n z Among them, n x 、n y and n z The maximum number of grids in the horizontal, vertical and height dimensions respectively; Range x 、Range y and Range z are the spatial spans of the horizontal, vertical and height dimensions respectively; max x and min x Respectively represent the maximum and minimum coordinate positions of the horizontal dimension, max y and min y Respectively represent the maximum and minimum coordinate positions of the longitudinal dimension; max z and min z Respectively represent the maximum and minimum coordinate positions of the height dimension; F is the minimum number of point cloud data contained in each grid; Indicates rounding up; Get the center of each subcluster in each class, and get the number of subclusters of other classes in the ε neighborhood of each subcluster center represents the number of subclusters that do not belong to the i-th category in the ε neighborhood of the j-th subcluster in the i-th category, ε = max(n x ,n y ,n z ); The variance of the eigenvector in each sub-cluster is used as the chaotic eigenvalue of the degree of chaos of the point cloud data in the corresponding sub-cluster based on and The product of is proportional to the number of base grids N T , get the number of triangle meshes corresponding to each subcluster: in, is the number of triangle meshes corresponding to the jth subcluster in the i-th class; α is the adjustment coefficient.

3. The three-dimensional visual modeling method of highway slope according to claim 1, characterized in that: The points are divided into several different classes according to the similarity of the feature vectors, specifically including: The feature vectors are clustered using density-based spatial clustering algorithms and k-means clustering algorithms, and corresponding category labels are added to each point according to the clustering results: Among them, p i is the i-th point in the highway slope point cloud data; It is a density-based spatial clustering algorithm for p i The first category label added, k is the total number of class labels; is the k-means clustering algorithm for p i Added second category label, Points with the same first and second category labels are classified into one category, so as to reallocate category labels to each point, and obtain several different categories based on the reallocated category labels.

4. The three-dimensional visual modeling method of highway slope according to claim 1, characterized in that: The extracting of the feature vector of each point based on the three-dimensional coordinates of each point and the signal reflection feature value specifically includes: For any point p in the point cloud dataset i (x i ,y i ,z i ), extract its signal reflection characteristic value θ i and the point set N(p i ): in, It's point p i d points in the neighborhood δ; Use local least squares fitting to fit each point p i and its neighborhood N(p i ) and take the quadratic surface as point p i The corresponding quadratic curvature; Point p i Neighborhood N(p i ) to obtain the corresponding quadratic surface curvature N(k i ): in, They are point p i The curvature of the quadratic surface corresponding to the d points in the neighborhood δ; θ i , k i and N(k i ) to obtain any point p i The eigenvector P i ={θ i ,k i ,N(k i )}.

5. A three-dimensional visual modeling device for highway slopes, characterized in that: include: A data acquisition module, used to collect point cloud data of the highway slope, wherein the point cloud data includes the three-dimensional coordinates of each point and the signal reflection characteristic value of each point; A clustering module is used to extract the feature vector of each point based on the three-dimensional coordinates and signal reflection feature values ​​of each point, and to classify each point into several different classes according to the similarity of the feature vectors. Each class includes several sub-clusters dispersed in geographical space. A triangular mesh quantity acquisition module is used to obtain a chaotic characteristic value used to characterize the degree of chaos of point cloud data in each subcluster and the number of subclusters belonging to other categories in the neighborhood of each subcluster; For each subcluster, the number of triangle meshes is set proportional to the chaotic eigenvalue and the number of subclusters belonging to other categories; The modeling module is used to connect the midpoints of each sub-cluster into non-overlapping triangular meshes using a triangulation method in combination with the set number of triangular meshes to obtain a three-dimensional model of the highway slope.

6. A computer-readable storage medium, characterized in that: The storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 4 is implemented.

7. A computer device, characterized in that: The method comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the method described in any one of claims 1 to 4 is implemented.

Images