CN-121999170-A - Coal mine deep roadway surface reconstruction method based on triangular prism voxels and spatial positioning

CN121999170ACN 121999170 ACN121999170 ACN 121999170ACN-121999170-A

Abstract

The application relates to the technical field of digital modeling, and particularly discloses a coal mine deep roadway surface reconstruction method based on triangular prism voxels and space positioning, wherein an electronic device acquires surrounding rock point clouds of a target coal mine roadway section and divides the surrounding rock point clouds into a plurality of slices; and performing triangular prism voxelization downsampling on each slice to obtain triangular prism voxels of a joint roadway section morphology, distributing surrounding rock points in surrounding rock point clouds to the triangular prism voxels to determine non-empty voxels and empty voxels, constructing a first centroid list of the non-empty voxels, wherein the first centroid list is used for indicating the mapping relation between the serial numbers of the non-empty voxels and the centroid coordinates of the non-empty voxels, supplementing the centroids of the empty voxels based on the first centroid list to determine a second centroid list, and finally constructing a triangular grid model based on the second centroid list to obtain a target model of a target coal mine roadway section, so that the modeling precision and efficiency of a coal mine deep roadway can be effectively improved.

Inventors

MAO QINGHUA
QIN SONG
ZHOU TONG
WANG PENG
WAN XIANG
XUE XUSHENG
WANG CHUANWEI
Chai Jianquan

Assignees

西安科技大学

Dates

Publication Date: 20260508
Application Date: 20260410

Claims (10)

1. The method for reconstructing the surface of the deep tunnel of the coal mine based on the triangular prism voxels and the spatial positioning is characterized by comprising the following steps: Acquiring surrounding rock point clouds of a target coal mine roadway section, and dividing the surrounding rock point clouds into a plurality of slices; performing triangular prism voxelization downsampling on each slice to obtain triangular prism voxels, wherein the triangular prism voxels are attached to the tunnel section morphology of the target coal mine tunnel section; distributing surrounding rock points in the surrounding rock point cloud to the triangular prism voxels to determine non-empty voxels and empty voxels, wherein the non-empty voxels represent voxels containing the surrounding rock points, and the empty voxels represent voxels not containing the surrounding rock points; The method comprises the steps of constructing a first centroid list of the non-empty voxels, wherein the first centroid list is used for indicating the mapping relation between the serial numbers of the non-empty voxels and the centroid coordinates of the non-empty voxels; the centroid of the empty voxel is complemented based on the first centroid list to determine a second centroid list, wherein the second centroid list comprises a mapping relation between the number of the non-empty voxel and the centroid coordinates of the non-empty voxel and a mapping relation between the number of the empty voxel and the centroid coordinates of the empty voxel; And constructing a triangular mesh model based on the second centroid list to obtain a target model of the target coal mine roadway section.
2. The method for reconstructing the surface of the deep tunnel of the coal mine based on the triangular prism voxels and the spatial localization according to claim 1, wherein the performing the triangular prism voxelized downsampling to obtain the triangular prism voxels comprises: Constructing an axial bounding box of the slice in a section perpendicular to the tunneling direction, wherein the axial bounding box represents a minimum cube for wrapping the slice; Dividing four sides of the axial bounding box according to preset equal parts to obtain divided line segments, and determining dividing points of the axial bounding box; Connecting the dividing points with two end points of the divided line segments on each side to form a plurality of triangle bottom surfaces; And performing stretching operation on the triangular bottom surfaces along the opposite direction of the tunneling direction to obtain a plurality of triangular prism voxels, wherein the stretching thickness of the stretching operation is the same as the thickness of the slice.
3. The method for reconstructing a coal mine deep roadway surface based on triangular prism voxels and spatial localization according to claim 2, wherein the constructing the first centroid list of the non-empty voxels comprises: Assigning numbers to the non-empty voxels according to the slices, edges and segmented line segments corresponding to the non-empty voxels to obtain the numbers of the non-empty voxels, wherein the numbers of the non-empty voxels comprise slice serial numbers, edge serial numbers and line segment serial numbers; Determining an average value of three-dimensional coordinates of all surrounding rock points in the non-empty voxels as a centroid coordinate of the non-empty voxels; and establishing a mapping relation between the serial numbers of the non-empty voxels and the corresponding centroid coordinates to obtain the first centroid list.
4. A method of reconstructing a coal mine deep roadway surface based on triangular prism voxels and spatial localization as recited in claim 3 wherein said filling the centroid of the empty voxels based on the first centroid list to determine a second centroid list comprises: determining a target non-empty voxel corresponding to the empty voxel in the first centroid list based on a geometric feature region of the empty voxel in the roadway surrounding rock section; Determining centroid coordinates of the empty voxels based on centroid coordinates of the target non-empty voxels; And adding the mapping relation between the serial numbers of the empty voxels and the corresponding centroid coordinates of the empty voxels into the first centroid list to obtain the second centroid list.
5. The method for reconstructing the surface of the deep tunnel of the coal mine based on the triangular prism voxels and the spatial localization according to claim 4, wherein the determining the target non-empty voxels corresponding to the empty voxels in the first centroid list based on the geometric feature region of the empty voxels in the tunnel surrounding rock section comprises: Determining whether first non-empty voxels which are larger than or equal to a preset number and are identical to the slice sequence numbers and the edge sequence numbers of the empty voxels exist in the first centroid list under the condition that the geometric feature region is a linear feature region; If the first non-empty voxels with the preset number exist in the first centroid list, determining the target non-empty voxels based on the first non-empty voxels, otherwise, determining whether the second non-empty voxels with the same edge sequence number and line segment sequence number as the empty voxels with the preset number or not exist in the first centroid list; and if the second non-empty voxels with the preset number exist in the first centroid list, determining the target non-empty voxels based on the second non-empty voxels, otherwise, determining the target non-empty voxels based on the third non-empty voxels with the symmetrical positions of the empty voxels in the first centroid list.
6. The method for reconstructing the surface of the deep tunnel of the coal mine based on the triangular prism voxels and the spatial localization according to claim 5, wherein the determining the target non-empty voxels corresponding to the empty voxels in the first centroid list based on the geometric feature region of the empty voxels in the tunnel surrounding rock section comprises: determining whether second non-empty voxels greater than or equal to a preset number exist in the first centroid list under the condition that the geometric feature region is a curve feature region; and if the second non-empty voxels with the preset number exist in the first centroid list, determining the target non-empty voxels based on the second non-empty voxels, otherwise, determining the target non-empty voxels based on the third non-empty voxels with the symmetrical positions of the empty voxels in the first centroid list.
7. The method for reconstructing the surface of the deep tunnel of the coal mine based on the triangular prism voxels and the spatial localization according to claim 1, wherein the constructing the triangular mesh model based on the second centroid list to obtain the target model of the target coal mine tunnel section comprises the following steps: connecting adjacent voxel centroids on the basis of the second centroid list in the same slice to obtain a tunnel section contour line; Connecting centroids of voxels with the same edge sequence number and line segment sequence number in adjacent slices to obtain a plurality of space quadrilaterals; And dividing each space quadrangle into two triangular surfaces, and summarizing the connection relation between the vertex coordinates and the vertexes of all the triangular surfaces to obtain the target model.
8. The method for reconstructing a coal mine deep roadway surface based on triangular prism voxels and spatial localization according to claim 1, wherein the dividing the surrounding rock point cloud into a plurality of slices comprises: and dividing the surrounding rock point cloud into a plurality of slices with equal thickness along the tunneling direction according to the preset length.
9. The method for reconstructing the surface of the deep tunnel of the coal mine based on the triangular prism voxels and the spatial localization of the invention as set forth in claim 1, wherein the acquiring the surrounding rock point cloud of the target coal mine tunnel section comprises: Acquiring a first point cloud of the target coal mine roadway section; Performing semantic segmentation on the first point cloud by using the trained model to obtain a semantic tag; and eliminating points in the first point cloud, wherein the semantic tags are non-surrounding rocks, and determining the surrounding rock point cloud according to the remaining points in which the semantic tags are surrounding rocks.
10. An electronic device, comprising a processor, a memory, and a memory storing instructions executable by the processor, wherein the instructions, when executed by the processor, implement the method for reconstructing a coal mine deep roadway surface based on triangular prism voxels and spatial localization according to any one of claims 1 to 9.

Description

Coal mine deep roadway surface reconstruction method based on triangular prism voxels and spatial positioning Technical Field The application relates to the technical field of digital modeling, in particular to a coal mine deep roadway surface reconstruction method based on triangular prism voxels and space positioning. Background In the engineering construction and safety monitoring of coal mine tunnel, the deep tunnel is easily deformed and converged due to high ground stress, and the like, so that the three-dimensional model of the tunnel surface is quickly and accurately reconstructed and is very important for deformation monitoring and disaster early warning. The existing symmetrical-based complementation method is only suitable for small-scale deletion, the existing complementation method based on deep learning (DEEP LEARNING, DL) needs a large amount of labeled training data, has high calculation cost and weak generalization capability, and cannot realize the efficient complementation of the large-area point cloud deletion, and the third is that the surface reconstruction algorithm (such as poise reconstruction) from the point cloud to a grid model is limited by the point cloud precision, the effect of common voxel filtering and the large-area point cloud deletion, the effect of the large-area point cloud is easy to be lost, and the additional filtering algorithm can not be ensured when the surface of the reconstructed point cloud is easy to be lost. Therefore, the above problems limit the modeling accuracy and efficiency of the deep tunnel in the coal mine, and a coal mine tunnel surface reconstruction method which is adaptive to the tunnel geometry, can efficiently process the point cloud deficiency and has a concise reconstruction flow is needed to support the dynamic monitoring and the safety evaluation of the deep tunnel. Disclosure of Invention The application provides a coal mine deep tunnel surface reconstruction method based on triangular prism voxels and spatial positioning, which can effectively improve the accuracy and efficiency of coal mine deep tunnel surface reconstruction. In order to achieve the above object, the present application provides the following technical solutions: In a first aspect, an embodiment of the present application provides a method for reconstructing a surface of a deep tunnel of a coal mine based on triangular prism voxels and spatial localization, the method comprising: acquiring surrounding rock point clouds of a target coal mine roadway section, and dividing the surrounding rock point clouds into a plurality of slices; executing triangular prism voxelization downsampling on each slice to obtain triangular prism voxels, wherein the triangular prism voxels are attached to the tunnel section morphology of the target coal mine tunnel section; Distributing surrounding rock points in the surrounding rock point cloud to triangular prism voxels to determine non-empty voxels and empty voxels, wherein the non-empty voxels represent voxels containing the surrounding rock points, and the empty voxels represent voxels not containing the surrounding rock points; the method comprises the steps of constructing a first centroid list of non-empty voxels, wherein the first centroid list is used for indicating the mapping relation between the serial numbers of the non-empty voxels and the centroid coordinates of the non-empty voxels; The method comprises the steps of carrying out patch on centroids of empty voxels based on a first centroid list to determine a second centroid list, wherein the second centroid list comprises a mapping relation between the numbers of non-empty voxels and centroid coordinates of non-empty voxels and a mapping relation between the numbers of the empty voxels and the centroid coordinates of the empty voxels; And constructing a triangular mesh model based on the second centroid list to obtain a target model of the target coal mine roadway section. In some embodiments of the present application, performing triangular prism voxel downsampling to obtain triangular prism voxels comprises: Constructing an axial bounding box of the slice in a section perpendicular to the tunneling direction of the roadway, wherein the axial bounding box characterizes a minimum cube for wrapping the slice; dividing four sides of the axial bounding box according to preset equal parts to obtain divided line segments, and determining dividing points of the axial bounding box; connecting the dividing points with two end points of the line segment divided on each edge to form a plurality of triangle bottom surfaces; And performing stretching operation on the triangular bottom surfaces along the opposite direction of the tunneling direction to obtain a plurality of triangular prism voxels, wherein the stretching thickness of the stretching operation is the same as the thickness of the slice. In some embodiments of the present application, constructing a first list of centroids for