Rock mass information acquisition method based on three-dimensional laser scanning

Abstract

The invention relates to a rock mass information acquisition method based on three-dimensional laser scanning, and the method comprises the steps: employing a three-dimensional laser scanner to scan a to-be-measured tunnel face, and obtaining the point cloud data of the to-be-measured tunnel face; resampling the point cloud data of the to-be-measured tunnel face by using a moving least square method, and obtaining a point cloud model of the to-be-measured tunnel face by using a greedy projection triangulation method; performing rock mass discontinuous surface occurrence extraction on the to-be-detected tunnel face point cloud model to obtain a discontinuous surface of the to-be-detected tunnel face; performing trace extraction on the to-be-detected tunnel face point cloud model by adopting a rock mass trace identification algorithm to obtain a trace of the to-be-detected tunnel face; and calculating the discontinuous surface of the to-be-measured tunnel face by adopting a roughness calculation method to obtain the roughness of the to-be-measured tunnel face. According to the method, the tunnel face rock mass influenced by blasting can be effectively extracted.

Description

Technical Field

[0001] The present invention relates to the technical field of rock tunnel engineering, and particularly to a method for obtaining rock mass information based on three-dimensional laser scanning. Background Art

[0002] The construction of rock tunnel engineering is difficult to control due to the heterogeneity, discontinuity and uncertainty of the medium. Obtaining the structural information of the rock stratum has great guiding significance for construction. Affected by factors such as project cost, construction period, and terrain, the preliminary exploration of tunnel engineering (such as drilling and engineering surface geophysical prospecting) often fails to understand the geological information of the excavation area in detail. However, during the construction stage of the tunnel, as the tunnel is continuously excavated, complex geological bodies are exposed in the form of the tunnel face, and more geological data that cannot be obtained during the design stage can be obtained.

[0003] At present, three-dimensional laser scanning technology (Light Detection and Ranging, LiDAR) is used to survey rock tunnels. However, the number of rock mass point clouds obtained by laser scanning is huge. If the algorithm is complex, the calculation time will be very long. In addition, the algorithm for discontinuity surface recognition generally only applies to rock mass slopes with obvious exposed discontinuity surfaces. For the tunnel face rock mass affected by blasting, the recognition effect of the above algorithm may be poor. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a method for obtaining rock mass information based on three-dimensional laser scanning, which can effectively extract the tunnel face rock mass affected by blasting.

[0005] The technical solution adopted by the present invention to solve its technical problems is: to provide a method for obtaining rock mass information based on three-dimensional laser scanning, including the following steps:

[0006] Use a three-dimensional laser scanner to scan the tunnel face to be measured, and obtain the point cloud data of the tunnel face to be measured;

[0007] Use the moving least squares method to resample the point cloud data of the tunnel face to be measured, and use the greedy projection triangulation method to obtain the point cloud model of the tunnel face to be measured;

[0008] Extract the occurrence of the rock mass discontinuity surface from the point cloud model of the tunnel face to be measured, and obtain the discontinuity surface of the tunnel face to be measured;

[0009] Use the rock mass trace recognition algorithm to extract the traces from the point cloud model of the tunnel face to be measured, and obtain the traces of the tunnel face to be measured;

[0010] Use the roughness calculation method to calculate the discontinuity surface of the tunnel face to be measured, and obtain the roughness of the tunnel face to be measured.

[0011] Performing the extraction of the occurrence of the rock mass discontinuity surface on the point cloud model of the heading face to be measured to obtain the discontinuity surface of the heading face to be measured, specifically including:

[0012] Automatically grouping the triangular mesh data of the point cloud model of the heading face to be measured by using the K-means algorithm;

[0013] Dividing and optimizing the discontinuity surfaces within each group according to the adjacent relationship and the included angle between them;

[0014] Using the random sample consensus plane fitting algorithm to fit the segmented discontinuity surface point cloud to obtain the discontinuity surface of the heading face to be measured.

[0015] The specific process of automatically grouping the triangular mesh data of the point cloud model of the heading face to be measured by using the K-means algorithm is as follows: calculating the sample density of each triangular mesh data in the point cloud model of the heading face to be measured, extracting the triangular mesh data that meets the sample density expectation and storing it in the set D, setting the triangular mesh data with the largest sample density in the set D as the initial center point and removing it from the set D; then selecting the triangular mesh data with the largest average sample distance from the initial center point in the set as the second center point and also removing it from the set D; repeating this process until k center points are selected; then calculating the distance between each triangular mesh data and each center point, and assigning each triangular mesh data to the center point closest to it; the center points and the triangular mesh data assigned to them represent a cluster; once all the triangular mesh data are assigned, the center point of each cluster will be recalculated according to the existing triangular mesh data in the cluster; this process will be repeated continuously until the termination condition is met.

[0016] The specific process of dividing and optimizing the discontinuity surfaces within each group according to the adjacent relationship and the included angle between them is as follows: finding adjacent triangular patches for the discontinuity surfaces within each group, if the total number of triangular patches is less than the threshold, and when the triangular patch is adjacent to the triangular patches in other discontinuity surface groups and the included angle between them meets the merging requirement, then assigning the triangular patch to other discontinuity surface groups.

[0017] Performing the extraction of the trace on the point cloud model of the heading face to be measured by using the rock mass trace recognition algorithm to obtain the trace of the heading face to be measured, specifically including:

[0018] Using the feature extraction method based on the tensor voting theory to extract the initial feature points that make up the trace on the point cloud model of the heading face to be measured;

[0019] Grouping the initial feature points and continuously iteratively growing according to the grouping result by using the growth algorithm to search for all segments that make up the trace.

[0020] Connect the trace segments according to the distance and angle relationship between adjacent two trace segments to obtain multiple traces.

[0021] When grouping the initial feature points, two initial feature points are assigned to the same group when they simultaneously meet the following two conditions:

[0022] The two feature points are on one side of a triangular patch;

[0023] The included angle between the normal vectors of the two feature points is less than the angle threshold.

[0024] Before connecting the trace segments according to the distance and angle relationship between adjacent two trace segments to obtain multiple traces, it further includes judging whether the included angle between the direction of the trace segment and the main direction is greater than the rejection angle threshold. If it is greater, the trace segment is rejected.

[0025] When calculating the roughness of the discontinuous surface of the to-be-measured heading face by using the roughness calculation method, specifically including:

[0026] Establish a local coordinate system of the discontinuous surface of the to-be-measured heading face, where the x-axis of the local coordinate system is parallel to the strike and the y-axis is parallel to the dip;

[0027] Along the dip of the discontinuous surface of the to-be-measured heading face, cut the discontinuous surface with a plane at different positions to obtain multiple strip-shaped point clouds;

[0028] Smooth the cut strip-shaped point clouds into contour lines, then sample at equal intervals Δ, and calculate the root mean square of different contour lines according to the sampling results, and convert the root mean square into the roughness of the to-be-measured heading face.

[0029] The calculation formula used when calculating the root mean square Z2 of different contour lines according to the sampling results is: where m is the number of strip-shaped point clouds, and z i is the vertical coordinate of the i-th point in the three-dimensional data of the discontinuous surface.

[0030] Advantageous Effects

[0031] Due to the adoption of the above technical solutions, compared with the prior art, the present invention has the following advantages and positive effects: The present invention realizes resampling, denoising and triangulation of point clouds by using the moving least squares method and greedy triangulation, effectively improves the geometric details of the point cloud model, and on this basis, extracts geometric information such as the occurrence distribution, roughness and traces of the discontinuous surface of the current heading face, so as to obtain accurate data of the heading face rock mass. Description of the Drawings

[0032] Figure 1It is a flowchart of the method for obtaining rock mass information based on 3D laser scanning according to the embodiments of the present invention. Specific embodiments

[0033] The present invention will be further described below in conjunction with specific embodiments. It should be understood that these embodiments are only used to illustrate the present invention and not to limit the scope of the present invention. In addition, it should be understood that after reading the content taught by the present invention, those skilled in the art can make various changes or modifications to the present invention, and these equivalent forms also fall within the scope defined by the appended claims of this application.

[0034] The embodiments of the present invention relate to a method for obtaining rock mass information based on 3D laser scanning, as Figure 1 shown, including the following steps:

[0035] Step 1, use a 3D laser scanner to scan the to-be-measured heading face to obtain the point cloud data of the to-be-measured heading face.

[0036] In this embodiment, the 3D laser scanner used is the new generation ultra-high-speed laser scanner Scanstation P30. The angular measurement accuracy of this 3D laser scanner is 8", and the ranging accuracy is 1.2mm + 10ppm, ensuring accurate and reliable results. The scanning rate of the scanner is as high as 1,000,000 points / second. The ultra-high-speed scanning can reduce the fieldwork time and save costs. At the same time, the scanning resolution can be flexibly selected according to user needs, which is suitable for construction environments with complex and changeable conditions.

[0037] Step 2, use the moving least squares method to resample the point cloud data of the to-be-measured heading face, and use the greedy projection triangulation method to obtain the point cloud model of the to-be-measured heading face.

[0038] In order to establish a complete model, it is necessary to smooth the surface and repair holes. In this embodiment, based on the point cloud data of the to-be-measured heading face, the moving least squares method (Moving Least Squares, MLS) is used to solve the problem of data resampling. The resampling algorithm reconstructs the missing part of the surface by performing high-order polynomial interpolation on the surrounding data points, which can effectively improve the geometric details of the point cloud model.

[0039] When studying the point cloud data of the heading face, to convert its discrete points into a continuous surface, triangulation must be carried out first to form small interconnected triangular planes, and then these small planes are aggregated on this basis to extract the geometric information of the heading face. In this embodiment, the greedy projection triangulation method is adopted, specifically: project the three-dimensional point cloud onto a certain plane, and then triangulate the projected point cloud within the plane to obtain the connection relationship of each point. In the process of triangulating the plane area in this embodiment, the Delaunay-based spatial region growth algorithm is used. This method forms a complete triangular mesh surface by selecting a sample triangular patch as the initial surface and continuously expanding the surface boundary. Finally, the topological connection between the original three-dimensional points is determined according to the connection relationship of the projected point cloud, and the obtained triangular mesh is the reconstructed surface model. The greedy projection triangulation method of this embodiment is based on the incremental surface growth law, starts with the creation of a starting triangle, and continuously adds new triangles until all the points in the point cloud are included or there are no more effective triangles.

[0040] Step 3: Extract the attitude of the rock mass discontinuity surface from the measured heading face point cloud model to obtain the discontinuity surface of the measured heading face.

[0041] The attitude of the rock mass discontinuity surface includes three elements: dip direction, dip angle, and strike. The extraction of the discontinuity surface attitude in this embodiment includes the following three parts:

[0042] (1) Automatic grouping of discontinuity surfaces

[0043] In this embodiment, an improved K-means algorithm based on sample density and validity index analysis is used to automatically group the triangular mesh data. Calculate the sample density of each triangular mesh data in the measured heading face point cloud model, extract the triangular mesh data that meets the sample density expectation and store it in set D, set the triangular mesh data with the largest sample density in set D as the initial center point and remove it from set D; then select the triangular mesh data with the largest average sample distance from the initial center point in the set as the second center point, and also remove it from set D; repeat this process until K center points are selected. Then calculate the distance between each triangular mesh data and each center point, and assign each triangular mesh data to the center point closest to it. The center points and the triangular mesh data assigned to them represent a cluster. Once all objects are assigned, the center point of each cluster will be recalculated according to the existing triangular mesh data in the cluster. This process will be repeated continuously until a certain termination condition is met. The termination condition can be that no (or the minimum number) of triangular mesh data is reassigned to different clusters, no (or the minimum number) of cluster centers change anymore, and the sum of squared errors is locally minimized.

[0044] (2) Discontinuity surface segmentation and optimization

[0045] Each set of discontinuous surfaces is segmented and optimized according to the adjacent relationship and the included angle between them. In order to facilitate the subsequent automatic fitting and calculation of the attitude of each discontinuous surface, it is necessary to segment the grouped discontinuous surfaces according to their adjacent relationship, so that the small triangular patches of each discontinuous surface after segmentation form a set independently.

[0046] The segmentation of the discontinuous surface is achieved by continuously finding adjacent triangular patches and incorporating them into the set until there are no adjacent triangular patches that can be incorporated. Adjacent triangular patches are defined as two triangular patches sharing two vertices. Note that the surface of the discontinuous surface is not a complete plane, and there are local unevenness. If grouped directly according to the unit normal vector, many small groups will be generated. Therefore, in this embodiment, it is necessary to optimize the segmented discontinuous surface again. The optimization principle is: for the discontinuous surfaces within each group, find adjacent triangular patches. If the total number of triangular patches is less than a given threshold, then judge the triangular patches in this group of discontinuous surfaces one by one, and allocate them to other groups of discontinuous surfaces according to the adjacent relationship and included angle with the triangular patches in other groups of discontinuous surfaces (the triangular patch and the triangular patch in other groups of discontinuous surfaces are in an adjacent relationship and the included angle between them meets the merging requirement, and this merging requirement can be that the included angle is less than a certain set threshold). After such smoothing processing, the problem of excessive fragmentation of the discontinuous surfaces of the tunnel face rock mass can be partially solved.

[0047] (3) Plane fitting of discontinuous surfaces

[0048] In this embodiment, the random sample consensus plane fitting algorithm is used to fit the point cloud of the segmented discontinuous surfaces. The discontinuous surfaces of the rock mass are usually uneven, with undulation and roughness, which reflect the surface unevenness of the discontinuous surface relative to the average plane. Larger undulation may affect the measurement results of the local attitude of the discontinuous surface. Therefore, the attitude can be calculated by calculating the fitting plane of the discontinuous surface. Common plane fitting methods such as the least squares method and the eigenvalue method can consider the error factors of the point cloud data, but cannot eliminate the abnormal point cloud with large fluctuations. In this embodiment, the random sample consensus algorithm (Random Sample Consensus, RANSAC) is used for three-dimensional point cloud plane fitting, which can consider the fluctuations caused by the unevenness of the discontinuous surface and obtain a better fitting plane. These fitting planes are the discontinuous surfaces of the tunnel face to be measured.

[0049] Step 4: Use the rock mass trace recognition algorithm to extract the traces from the point cloud model of the tunnel face to be measured, and obtain the traces of the tunnel face to be measured.

[0050] Based on the characteristics that the edge lines are manifested as sharp edge points and corner points on the triangular mesh surface in this embodiment, a three-dimensional rock mass trace recognition algorithm is proposed. First, a robust algorithm for feature extraction based on the tensor voting theory is used to extract the initial feature points forming the traces on the triangular mesh. Secondly, four-step optimization processing techniques of "grouping", "growing algorithm to extract trace segments", "connecting trace segments" and "linearizing traces" are adopted to solve problems such as the short traces extracted due to the uneven surface of the rock mass, and the extracted traces are more continuous and penetrating, conforming to the actual discontinuous surface conditions. Specifically as follows:

[0051] In the point cloud model of the tunnel face to be measured, the vertex v of each triangular patch i ∈V can be represented in the Cartesian coordinate system as v i =(v ix , v iy , v iz ). Its tensor voting matrix can be expressed as: Where n fi =(a, b, c) T is the unit normal vector of the triangular patch fi, and N f (v) represents the set composed of all triangular patches within the 1-ring neighborhood of the vertex v i , and μ fi is the weight function, expressed as: Where A(f i ) is the area of the triangular patch fi, A max is the largest area among all triangular patches, c fi is the centroid of the triangular patch fi, and σ is the side length of the minimum cubic bounding box of the triangles within the 1-ring neighborhood of the vertex v i .

[0052] The above-mentioned normal-based tensor voting T v is a symmetric positive semi-definite tensor. Therefore, after diagonalizing it, three eigenvalues λ1, λ2, and λ3 are obtained. According to the characteristics of the three eigenvalues, the vertices on the triangular mesh can be divided into points on the surface, points on the edge, and points on the corner.

[0053] Through the above analysis, all the feature points forming the traces can be extracted. However, these feature points are all in one group and further analysis is needed to store the points on each trace separately as a group. In this embodiment, when two feature points simultaneously meet the following two conditions, they can be assigned to the same group:

[0054] (1) The two feature points are on one edge of a triangular patch;

[0055] (2) The included angle between the normal vectors of the two feature points is less than the angle threshold θ1.

[0056] Through the above grouping process, the feature points can be divided into different groups, and each group is composed of adjacent feature points.

[0057] Based on the idea that the trace is composed of linear feature points, this embodiment uses a growth algorithm to continuously iterate and grow to search for all segments that make up the trace. First, the initial principal direction of the entire point cloud is calculated by principal component analysis Note that in the initial stage of trace growth, due to the undulation and unevenness of the rock mass surface, the initial principal direction is uniformly used as the current calculation principal direction When the number of points in the point set {T V} is greater than 10, the initial principal direction is calculated The principal direction obtained by performing principal component analysis on five points adjacent to the growing point in the point set is

[0058] In the selection of the seed point, it should be avoided that the growth direction starts from the branch of the trace, so that the generated trace is too short. The strategy adopted here is to orthogonally project all points onto the principal direction and select the point farthest from the randomly selected point as the seed point to ensure that the selected seed point is at the end point of the longest trace.

[0059] The growth of feature points needs to satisfy one of the following two conditions:

[0060]

[0061] num(V N ) = 1.

[0062] In the formula, the vector points from the current growing point V to the neighborhood feature point V N , is the principal direction and the vector is the included angle, θ2 is the user-defined angle threshold. num(V N ) = 1 is used to prevent the growth from being interrupted due to singular triangular patches, and num(V N ) is the number of neighborhood feature points V N .

[0063] Each trace segment consists of a list of vertices and can be described as (X st , Y st , Z st , X end , Y end , Z end , l, m, n), where (X st , Y st , Z st ), (Xend ,Y end ,Z end ) and (l,m,n) are the starting point, midpoint and direction of the trace segment respectively. The connection criterion is defined by the angle threshold θ3 and the distance threshold d between the two trace segments, for example, the angle between the two trace segments is less than the angle threshold θ3, and the distance between the two trace segments is less than the distance threshold d.

[0064] After the trace segments are connected, the surface of the rock discontinuity surface is irregular, containing small-scale roughness and large-scale undulation. Therefore, the direction of the trace end changes drastically relative to the main direction. On the other hand, the growth algorithm used will also generate some redundant trace segments. In order to obtain a continuous and smooth trace, the connected trace needs to be smoothed. Therefore, an angle threshold θ4 can be set to remove these redundant trace segments, and the trace segments with an angle greater than θ4 with the main direction can be deleted. Then, the trace segments are connected, and finally the obtained trace is smoothed.

[0065] Step 5: Calculate the discontinuous surface of the tunnel face to be measured by using a roughness calculation method to obtain the roughness of the tunnel face to be measured.

[0066] In this embodiment, based on the discontinuous surfaces of the tunnel face to be measured obtained in step 3, the roughness of each discontinuous surface is extracted using the following three steps:

[0067] (1) establishing a local coordinate system of the discontinuity surface of the tunnel face to be measured, wherein the x-axis of the local coordinate system is parallel to the strike, and the y-axis is parallel to the dip;

[0068] (2) along the inclination of the discontinuity surface of the tunnel face to be measured, cutting the discontinuity surface with a plane at different positions to obtain a plurality of strip-shaped point clouds;

[0069] (3) Smooth the cut strip point cloud into contour lines, then sample at equal intervals Δ, and calculate the root mean square of different contour lines based on the sampling results, and convert the root mean square into the roughness of the measured face. The root mean square Z2 is calculated as follows: Among them, m is the number of strip point clouds, z i is the vertical coordinate of the i-th point in the three-dimensional data of the discontinuity surface. The conversion formula between the roughness JRC of the measured tunnel face and the root mean square Z2 is: JRC = 32.2 + 32.47logZ2.

[0070] It is not difficult to find that the present invention realizes resampling, denoising and triangulation of point clouds by using moving least squares and greedy triangulation, effectively improves the geometric details of the point cloud model, and extracts geometric information such as the occurrence distribution, roughness and trace of the discontinuous surface of the current heading face on this basis, so as to obtain accurate data of the heading face rock mass.

Claims

1. A method for obtaining rock mass information based on three-dimensional laser scanning, characterized in that: The following steps are involved: A 3D laser scanner is used to scan the tunnel face to be measured to obtain point cloud data of the tunnel face to be measured; The point cloud data of the tunnel face to be measured are resampled by using a moving least squares method, and a point cloud model of the tunnel face to be measured is obtained by using a greedy projection triangulation method; Extracting the discontinuity of the rock mass from the point cloud model of the tunnel face to be measured to obtain the discontinuity of the tunnel face to be measured; Using a rock mass trace recognition algorithm to extract traces from a point cloud model of a tunnel face to be measured, to obtain traces of the tunnel face to be measured; The roughness calculation method is used to calculate the discontinuous surface of the tunnel face to be measured to obtain the roughness of the tunnel face to be measured.

2. The rock mass information acquisition method based on three-dimensional laser scanning according to claim 1 is characterized in that: The step of extracting the discontinuity of the rock mass from the point cloud model of the tunnel face to be measured to obtain the discontinuity of the tunnel face to be measured specifically includes: Using K-means algorithm to automatically group the triangulated network data of the point cloud model of the tunnel face to be measured; The discontinuous surfaces in each group are divided and optimized according to the adjacent relationship and the angle between them; The segmented discontinuity surface point cloud is fitted using a random sampling consistency plane fitting algorithm to obtain the discontinuity surface of the tunnel face to be measured.

3. The rock mass information acquisition method based on three-dimensional laser scanning according to claim 2 is characterized in that: The K-means algorithm is used to automatically group the triangulated network data of the measured tunnel face point cloud model, specifically: the sample density of each triangulated network data in the measured tunnel face point cloud model is calculated, the triangulated network data that meets the sample density expectation is extracted and stored in a set D, the triangulated network data with the largest sample density in the set D is set as the initial center point and is removed from the set D; the triangulated network data with the largest average sample distance from the initial center point is selected from the set as the second center point, and it is also removed from the set D; this process is repeated until k center points are selected; then the distance between each triangulated network data and each center point is calculated, and each triangulated network data is assigned to the center point closest to it; the center and the triangulated network data assigned to them represent a cluster; once all triangulated network data are assigned, the center point of each cluster will be recalculated based on the existing triangulated network data in the cluster; This process will be repeated until the termination condition is met.

4. The method for obtaining rock mass information based on three-dimensional laser scanning according to claim 2, characterized in that: The discontinuous faces in each group are divided and optimized according to the adjacent relationship and the angle between them, specifically: adjacent triangular face patches are searched for the discontinuous faces in each group, if the total number of triangular face patches is less than a threshold value, and when the triangular face patch is adjacent to the triangular face patches in other discontinuous face groups and the angle between the two meets the merging requirements, the triangular face patch is allocated to other discontinuous face groups.

5. The method for obtaining rock mass information based on three-dimensional laser scanning according to claim 1, characterized in that: The method of extracting traces from the point cloud model of the tunnel face to be measured by using a rock mass trace recognition algorithm to obtain the traces of the tunnel face to be measured specifically includes: The feature extraction method based on tensor voting theory is used to extract the initial feature points that constitute the traces from the point cloud model of the tested tunnel face. The initial feature points are grouped, and a growth algorithm is used to iteratively grow the points according to the grouping results to search for all the segments constituting the trace; The trace segments are connected according to the distance and angle relationship between two adjacent trace segments to obtain multiple traces.

6. The method for obtaining rock mass information based on three-dimensional laser scanning according to claim 5, characterized in that: When the initial feature points are grouped, two initial feature points are assigned to the same group when they simultaneously meet the following two conditions: Two feature points are on an edge of a triangular patch; The angle between the normal vectors of two feature points is less than the angle threshold.

7. The method for obtaining rock mass information based on three-dimensional laser scanning according to claim 5, characterized in that: Before the trace segments are connected according to the distance and angle relationship between two adjacent trace segments to obtain multiple traces, the method further includes determining whether the angle between the trace segment direction and the main direction is greater than a rejection angle threshold, and if so, rejecting the trace segment.

8. The method for obtaining rock mass information based on three-dimensional laser scanning according to claim 1, characterized in that: The method of calculating the discontinuous surface of the tunnel face to be measured by using the roughness calculation method to obtain the roughness of the tunnel face to be measured specifically includes: Establishing a local coordinate system of the discontinuity surface of the tunnel face to be measured, wherein the x-axis of the local coordinate system is parallel to the strike, and the y-axis is parallel to the dip; Along the inclination of the discontinuous surface of the tunnel face to be measured, the discontinuous surface is cut with a plane at different positions to obtain a plurality of strip-shaped point clouds; The strip point cloud obtained by cutting is smoothed into contour lines, and then sampled at equal intervals Δ. The root mean square of different contour lines is calculated based on the sampling results, and the root mean square is converted into the roughness of the measured face.

9. The method for obtaining rock mass information based on three-dimensional laser scanning according to claim 8, characterized in that: The calculation formula used when calculating the root mean square Z2 of different contour lines according to the sampling results is: Among them, m is the number of strip point clouds, z i is the vertical coordinate of the i-th point in the three-dimensional data of the discontinuity surface.

Images