Method and system for extracting geometric fixed points at intersections of underground tunnels

Abstract

The present invention provides a method and system for extracting geometric fixed points at underground tunnel intersections, comprising: step S1: inputting a three-dimensional point cloud of the intersection to be processed to obtain a fitted ground plane equation and mine tunnel height information; based on the ground plane equation and the mine tunnel height information, segmenting the three-dimensional point cloud of the intersection to be processed into ground points and non-ground points, and extracting the most structurally distinct and representative wall points from the non-ground points; step S2: projecting the wall points from a laser radar coordinate system to a ground plane coordinate system to obtain a transformed wall point cloud; and step S3: converting the transformed wall point cloud into a two-dimensional occupancy grid image, and extracting the geometric fixed points of the intersection using a Delaunay triangulation method. This method can overcome the problem of significant changes in the structural features of the point cloud of the same intersection scanned by the laser radar due to large changes in the position of the laser radar during flight.

Description

Technical Field

[0001] The present invention relates to the technical field of three-dimensional point cloud data processing, and in particular to a method and system for extracting geometric fixed points at intersections of underground tunnels. Background Art

[0002] Drones offer advantages such as lightness, flexibility, excellent maneuverability, and fast inspection speeds. Equipped with various environmental sensing devices such as lidar and visual sensors, they can replace workers in performing dangerous tasks such as inspecting complex mine environments. This significantly reduces the incidence of coal mine accidents, improves production efficiency, and promotes the mechanization, automation, informationization, and intelligent development of the coal mining industry. Underground coal mines are characterized by narrow tunnels, simple geometric features, a lack of GPS signals, and low ambient illumination. This makes existing mapping and positioning technologies used on the ground inapplicable in coal mine environments, making autonomous positioning of underground drone inspections difficult and risky. In degraded underground environments with simple tunnel geometry and high wall feature similarity, intersections, as key topological features, provide crucial semantic information for global relocalization of drones in the underground environment, and are crucial for accurate positioning and autonomous navigation of drones. Summary of the Invention

[0003] In view of the defects in the prior art, the present invention provides a method and system for extracting geometric fixed points at intersections of underground tunnels.

[0004] According to the present invention, a method and system for extracting geometric fixed points at intersections of underground tunnels are provided, and the scheme is as follows:

[0005] In a first aspect, a method for extracting geometric fixed points at an intersection of an underground tunnel is provided, the method comprising:

[0006] Step S1: Input a 3D point cloud of the intersection to be processed, and obtain a fitted ground plane equation and mine tunnel height information based on the RANSAC plane fitting algorithm; based on the ground plane equation and the mine tunnel height information, segment the 3D point cloud of the intersection to be processed into ground points and non-ground points, and extract the most structurally distinct and representative wall points from the non-ground points;

[0007] Step S2: Projecting the wall points from the laser radar coordinate system to the ground plane coordinate system to obtain a wall point cloud after coordinate transformation;

[0008] Step S3: Convert the wall point cloud after coordinate conversion into a two-dimensional occupancy grid image, and use the Delaunay triangulation method to extract the geometric fixed points of the intersection.

[0009] Preferably, the step S1 includes:

[0010] Input the 3D point cloud of the intersection to be processed L P, the three-dimensional point cloud L The points in P are sorted by z value, and Nseed points with the smallest z value are selected as seed points to estimate the initial plane equation G0 of the ground;

[0011] Set the plane fitting distance threshold ε p , calculate the 3D point cloud L The calculated distance from all points in P to the initial plane equation G0 is selected to be less than the threshold ε p The point is used as a new seed point to update the initial plane equation G0; repeat N iterations g After times, the final fitting ground plane G is obtained plane ;

[0012] According to the fitted ground plane equation G plane , original 3D point cloud L P is segmented into non-ground points L P ng and ground points L P g ; Calculate non-ground points L P ng Distance d from the ground g , minus the ground distance d g >Height information of mine tunnel h mine Points, thereby segmenting the wall points L P w .

[0013] Preferably, step S3 includes:

[0014] Step S3.1: Project the wall point cloud onto the ground plane G P w Dimensionality reduction to X g O G Y G plane, converted to a binary image with a resolution of δ×δ, and retaining k×km 2 The square area is taken as the region of interest, where the origin O G is the center of the square area, and the square area is discretized into A binary image of

[0015] Step S3.2: Use binary unification filtering to remove the smaller isolated areas in the binary image and extract the pixel positions of the remaining continuous areas to form a two-dimensional point set P 2d ;

[0016] Step S3.3: For the two-dimensional point set P 2d Perform Delaunay triangulation and calculate the local minimum width d of the lane based on the triangulation resultsw .

[0017] Preferably, the step S3.3 includes:

[0018] The triangles outside the wall outline are all unstable obtuse triangle features. All unstable obtuse triangles are removed; the remaining acute triangles are all located on the inside of the tunnel wall. The local minimum width d of the tunnel is estimated by calculating the height of the shortest side of each acute triangle. w ;

[0019] The length of each acute triangle is equal to the local minimum width d of the roadway. w Compare, if any side of the triangle is less than d w , then remove the triangle;

[0020] After removing all edges less than d w After the triangle is formed, the long corridor has no three sides greater than the local minimum width d of the roadway. w The triangle is retained, and the T-intersection has only one triangle with three sides greater than the local minimum width d of the roadway. w The triangle at the intersection has only two sides whose three sides are greater than the local minimum width d of the roadway. w The triangle is retained for the calculation of invariant points;

[0021] The geometrically invariant point of a T-shaped intersection is located at the center of the circumscribed circle of the acute triangle. The midpoint of the line connecting the centers of the circumscribed circles of the two triangles of a cross-shaped intersection is considered to be the location of the geometrically invariant point of the cross-shaped intersection. Similarly, for intersections containing N w The star-shaped intersection of the walls is eliminated by removing obtuse triangles and triangles whose shortest side is less than the local minimum width d of the lane. w After the acute triangle, only N w -2 stable acute triangles remain. The common feature of these stable acute triangles is that their three vertices are located on different walls. w -The geometric mean of the centers of the circumscribed circles of the two acute triangles is used as the geometric fixed point of the star intersection.

[0022] In a second aspect, a system for extracting geometric fixed points at intersections of underground tunnels is provided, the system comprising:

[0023] Model M1: Inputs a 3D point cloud of the intersection to be processed and uses the RANSAC plane fitting algorithm to obtain the fitted ground plane equation and mine roadway height information. Based on the ground plane equation and mine roadway height information, the 3D point cloud of the intersection to be processed is segmented into ground points and non-ground points, and the most structurally distinct and representative wall points are extracted from the non-ground points.

[0024] Model M2: Project the wall points from the LiDAR coordinate system to the ground plane coordinate system to obtain a wall point cloud after coordinate transformation;

[0025] Model M3: The wall point cloud after coordinate conversion is converted into a two-dimensional occupancy grid image, and the Delaunay triangulation method is used to extract the geometric fixed points of the intersection.

[0026] Preferably, the module M1 includes:

[0027] Input the 3D point cloud of the intersection to be processed L P, the three-dimensional point cloud L The points in P are sorted by z value, and Nseed points with the smallest z value are selected as seed points to estimate the initial plane equation G0 of the ground;

[0028] Set the plane fitting distance threshold ε p , calculate the 3D point cloud L The calculated distance from all points in P to the initial plane equation g0 is selected to be less than the threshold ε p The point is used as a new seed point to update the initial plane equation G0; repeat N iterations g After times, the final fitting ground plane G is obtained plane ;

[0029] According to the fitted ground plane equation G plane , original 3D point cloud L P is segmented into non-ground points L P ng and ground points L P g ; Calculate non-ground points L P ng Distance d from the ground g , minus the ground distance d g >Height information of mine tunnel h mine Points, thereby segmenting the wall points L P w .

[0030] Preferably, the module M3 includes:

[0031] Module M3.1: Wall point cloud projected onto the ground plane G P w Dimensionality reduction to X G O G Y G plane, converted to a binary image with a resolution of δ×δ, and retaining k×km 2 The square area is taken as the region of interest, where the origin O G is the center of the square area, and the square area is discretized into A binary image of

[0032] Module M3.2: Use binary unification filtering to remove smaller isolated areas in the binary image and extract the pixel positions of the remaining continuous areas to form a two-dimensional point set P 2d ;

[0033] Module M3.3: For a two-dimensional point set P 2d Perform Delaunay triangulation and calculate the local minimum width d of the lane based on the triangulation results w .

[0034] Preferably, the module M3.3 includes:

[0035] The triangles outside the wall outline are all unstable obtuse triangle features. All unstable obtuse triangles are removed; the remaining acute triangles are all located on the inside of the tunnel wall. The local minimum width d of the tunnel is estimated by calculating the height of the shortest side of each acute triangle. w ;

[0036] The length of each acute triangle is equal to the local minimum width d of the roadway. w Compare, if any side of the triangle is less than d w , then remove the triangle;

[0037] After removing all edges less than d w After the triangle is formed, the long corridor has no three sides greater than the local minimum width d of the roadway. w The triangle is retained, and the T-intersection has only one triangle with three sides greater than the local minimum width d of the roadway. w The triangle at the intersection has only two sides whose three sides are greater than the local minimum width d of the roadway. w The triangle is retained for the calculation of invariant points;

[0038] The geometrically invariant point of a T-shaped intersection is located at the center of the circumscribed circle of the acute triangle. The midpoint of the line connecting the centers of the circumscribed circles of the two triangles of a cross-shaped intersection is considered to be the location of the geometrically invariant point of the cross-shaped intersection. Similarly, for intersections containing N w The star-shaped intersection of the walls is eliminated by removing obtuse triangles and triangles whose shortest side is less than the local minimum width d of the lane. w After the acute triangle, only N w -2 stable acute triangles remain. The common feature of these stable acute triangles is that their three vertices are located on different walls. w -The geometric mean of the centers of the circumscribed circles of the two acute triangles is used as the geometric fixed point of the star intersection.

[0039] In a third aspect, a computer-readable storage medium storing a computer program is provided, and when the computer program is executed by a processor, the steps in the method for extracting geometric fixed points at intersections of underground tunnels are implemented.

[0040] In a fourth aspect, an electronic device is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program, when executed by the processor, implements the steps in the method for extracting geometric fixed points at intersections of underground tunnels.

[0041] Compared with the prior art, the present invention has the following beneficial effects:

[0042] The present invention can be used for degraded underground tunnel environments with a single geometric structure and high similarity of wall features. It utilizes the inherent characteristics of the irregular intersection geometry to extract the center of the intersection located on the ground of the underground tunnel as a fixed viewpoint, thereby solving the problem of large feature differences of the same intersection at different observation points.

[0043] Other beneficial effects of the present invention will be explained through the introduction of specific technical features and technical solutions in the specific implementation methods. Those skilled in the art should be able to understand the beneficial technical effects brought about by the introduction of these technical features and technical solutions. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Other features, objects and advantages of the present invention will become more apparent upon reading the detailed description of non-limiting embodiments with reference to the following drawings:

[0045] Figure 1 It is a schematic diagram of the process of the present invention;

[0046] Figure 2 Schematic diagram of the wall 3D point cloud segmentation method;

[0047] Figures 3(a) to 3(c) are schematic diagrams showing examples of the 3D point cloud segmentation results for a wall. Figure 3(a) shows the point cloud for a long corridor, Figure 3(b) shows the point cloud for a T-junction, and Figure 3(c) shows the point cloud for an intersection.

[0048] Figure 4 (a) ~ (b) shows the laser radar coordinate system O L -X L Y L Z L To the ground plane coordinate system O G -X G Y G Z G Schematic diagram of coordinate transformation;

[0049] Figures 5(a) to (l) are schematic diagrams of the detection process of invariant points in the underground semantic topological structure, where Figure 5(a) is the point cloud of the long corridor, Figure 5(b) is the binary image of the long corridor, Figure 5(c) is the triangulation of the long corridor, Figure 5(d) is the invariant points of the long corridor, Figure 5(e) is the three-dimensional point cloud of the T-junction, Figure 5(f) is the binary image of the T-junction, Figure 5(g) is the triangulation of the T-junction, Figure 5(h) is the invariant points of the T-junction, Figure 5(i) is the three-dimensional point cloud of the intersection, Figure 5(j) is the binary image of the intersection, Figure 5(k) is the triangulation of the intersection, and Figure 5(l) is the invariant points of the intersection. DETAILED DESCRIPTION

[0050] The present invention will be described in detail below with reference to specific embodiments. The following examples will help those skilled in the art to further understand the present invention, but are not intended to limit the present invention in any form. It should be noted that, for those skilled in the art, several changes and improvements can be made without departing from the scope of the present invention. These all fall within the scope of protection of the present invention.

[0051] The embodiment of the present invention provides a method for extracting geometric fixed points at intersections of underground tunnels. The method extracts stable ground plane features in the tunnel environment from the three-dimensional point cloud of the intersection through a plane fitting method. By analyzing the internal commonalities and external differences of different intersections, the key area features of the point cloud are extracted and the dimension is reduced to a two-dimensional occupancy grid image. The geometric fixed points of the intersection are extracted based on the Delaunay triangulation method. Figure 1 As shown, the method specifically includes the following contents:

[0052] Step S1: Input the 3D point cloud of the intersection to be processed L P, obtain the fitted ground plane equation G based on the RANSAC plane fitting algorithm plane and the height information of the mine tunnel h mine ; Based on the ground plane equation G plane and the height information of the mine tunnel h mine , the 3D point cloud of the intersection to be processed L P is segmented into ground points L P g and non-ground points L P ng , from non-ground points L P ng Extract the most structurally distinct and representative wall points L P w .

[0053] Reference Figure 2 As shown, input the 3D point cloud of the intersection to be processed L P, the three-dimensional point cloud LSort the points in P by their z values and select N with the smallest z value. seed Points are used as seed points to estimate the initial plane equation G0 of the ground. Set the plane fitting distance threshold ε p , calculate the 3D point cloud L The calculated distance from all points in P to the initial plane equation G0 is selected to be less than the threshold ε p The point is used as a new seed point to update the initial plane equation G0; repeat N iterations g After times, the final fitting ground plane G can be obtained plane ;

[0054] According to the fitted ground plane equation G plane , original 3D point cloud L P can be segmented into non-ground points L P ng and ground points L P g ; Since the irregular mine top surface does not meet the plane assumption, the non-ground points are calculated L P ng Distance d from the ground g , minus the ground distance d g >Height information of mine tunnel h mine Points, thereby segmenting the wall points L P w .

[0055] Schematic diagram of the wall 3D point cloud segmentation effect example Figure 3(a) 、 3(b) , as shown in 3(c). Figure 3(a) 、 3(b) ,3(c) are the point clouds of long corridor, T-intersection and cross intersection respectively.

[0056] Step S2: Point the wall L P w Project the laser radar coordinate system to the ground plane coordinate system to obtain the wall point cloud after coordinate transformation G P w .

[0057] like Figure 4(a) 、 4(b) As shown, the wall point cloud L P w From the laser radar coordinate system O L -X L Y L Z L Transform to the ground plane coordinate system O G -X G Y G Z G In the ground plane coordinate system OG -X G Y G Z G The wall point cloud in is represented as G P w .

[0058] Step S3: Convert the wall point cloud after coordinate conversion G P w The image is converted into a two-dimensional occupancy grid, and the Delaunay triangulation method is used to extract the geometric fixed points of the intersection.

[0059] As shown in FIG5 , step S3 includes:

[0060] Step S3.1: Project the image onto the ground plane. Figure 5(a) 、 5(e) , 5(i) shows the wall point cloud G P w Dimensionality reduction to X G O G Y G plane, converted to a binary image with a resolution of δ×δ, and retaining k×km 2 The square area is taken as the region of interest, such as Figure 5(b) 、 5(f) , 5(j). The origin O G is the center of the square area, and the square area is discretized into A binary image of .

[0061] Step S3.2: Use binary unification filtering to remove the smaller isolated areas in the binary image and extract the pixel positions of the remaining continuous areas to form a two-dimensional point set P 2d ;

[0062] Step S3.3: For the two-dimensional point set P 2d Perform Delaunay triangulation, Figure 5(c) 、 5(g) 5(k) is the result of triangulation. The local minimum width d of the lane is calculated based on the result of triangulation. w .

[0063] analyze Figure 5(c) 、 5(g) From the triangulation results shown in Figure 5(k), it can be seen that the triangles outside the wall contour are all unstable obtuse triangle features, so all unstable obtuse triangles in the figure are removed. Figure 5(d) 、 5(h) As shown in 5(i), the remaining acute triangles are all located on the inner side of the tunnel wall. The local minimum width d of the tunnel is estimated by calculating the height of the shortest side of each acute triangle. w .

[0064] The length of each acute triangle is equal to the local minimum width d of the roadway. w Compare, if any side of the triangle is less than d w , then remove the triangle. After removing all triangles with side lengths less than d w After the triangle, Figure 5(d) 、 5(h) As shown in Figure 5(l), the long corridor has no three sides larger than the local minimum width d of the lane. w The T-intersection has only one side whose three sides are greater than the local minimum width d of the roadway. w The acute triangle at the intersection has only two sides whose three sides are greater than the local minimum width d of the roadway. w acute triangle.

[0065] As shown in Figure 5(h), the geometric invariant point of the T-intersection is located at the center of the circumscribed circle of the acute triangle, that is, the point shown. As shown in Figure 5(l), the midpoint of the line connecting the centers of the circumscribed circles of the two acute triangles of the cross-shaped intersection is considered to be the location of the geometric invariant point of the cross-shaped intersection. Similarly, for the intersection containing N w The star-shaped intersection of the walls is eliminated by removing obtuse triangles and triangles whose shortest side is less than the local minimum width d of the lane. w After the acute triangle, only N w -2 stable acute triangles remain. The common feature of these stable acute triangles is that their three vertices are located on different walls. w -The geometric mean of the centers of the circumscribed circles of the two acute triangles is used as the geometric fixed point of the star intersection.

[0066] The present invention also provides a system for extracting geometric fixed points at underground roadway intersections. The system can be implemented by executing the process steps of the method for extracting geometric fixed points at underground roadway intersections. That is, those skilled in the art can understand the method as a preferred embodiment of the system for extracting geometric fixed points at underground roadway intersections. The specific contents of the system are as follows:

[0067] Module M1: Input the 3D point cloud of the intersection to be processed L P, obtain the fitted ground plane equation G based on the RANSAC plane fitting algorithm plane and the height information of the mine tunnel h mine ; Based on the ground plane equation G plane and the height information of the mine tunnel h mine , the 3D point cloud of the intersection to be processed L P is segmented into ground points L P g and non-ground pointsL P ng , from non-ground points L P ng Extract the most structurally distinct and representative wall points L P w .

[0068] Reference Figure 2 As shown, input the 3D point cloud of the intersection to be processed L P, the three-dimensional point cloud L Sort the points in P by their z values and select N with the smallest z value. seed Points are used as seed points to estimate the initial plane equation G0 of the ground. Set the plane fitting distance threshold ε p , calculate the 3D point cloud L The calculated distance from all points in P to the initial plane equation G0 is selected to be less than the threshold ε p The point is used as a new seed point to update the initial plane equation G0; repeat N iterations g After times, the final fitting ground plane G can be obtained plane ;

[0069] According to the fitted ground plane equation G plane , original 3D point cloud L P can be segmented into non-ground points L P ng and ground points L P g ; Since the irregular mine top surface does not meet the plane assumption, the non-ground points are calculated L P ng Distance d from the ground g , minus the ground distance d g >Height information of mine tunnel h mine Points, thereby segmenting the wall points L P w .

[0070] Schematic diagram of the wall 3D point cloud segmentation effect example Figure 3(a) 、 3(b) , as shown in 3(c). Figure 3(a) 、 3(b) ,3(c) are the point clouds of long corridor, T-intersection and cross intersection respectively.

[0071] Module M2: Point the wall L P w Project the laser radar coordinate system to the ground plane coordinate system to obtain the wall point cloud after coordinate transformation G P w .

[0072] like Figure 4(a) 、 4(b) As shown, the wall point cloud L P w From the laser radar coordinate system O L -X L Y L Z L Transform to the ground plane coordinate system O G -X G Y G Z G In the ground plane coordinate system O G -X G Y G Z G The wall point cloud in is represented as G P w .

[0073] Module M3: Convert the wall point cloud after coordinate conversion G P w The image is converted into a two-dimensional occupancy grid, and the Delaunay triangulation method is used to extract the geometric fixed points of the intersection.

[0074] As shown in Figure 5, module M3 includes:

[0075] Module M3.1: Projecting onto the ground plane Figure 5(a) 、 5(e) , 5(i) shows the wall point cloud G P w Dimensionality reduction to X G O G Y G plane, converted to a binary image with a resolution of δ×δ, and retaining k×km 2 The square area is taken as the region of interest, such as Figure 5(b) 、 5(f) , 5(j). The origin O G is the center of the square area, and the square area is discretized into A binary image of .

[0076] Module M3.2: Use binary unification filtering to remove smaller isolated areas in the binary image and extract the pixel positions of the remaining continuous areas to form a two-dimensional point set P 2d ;

[0077] Module M3.3: For a two-dimensional point set P 2d Perform Delaunay triangulation, Figure 5(c) 、 5(g) 5(k) is the result of triangulation. The local minimum width d of the lane is calculated based on the result of triangulation. w .

[0078] analyze Figure 5(c) 、 5(g) From the triangulation results shown in Figure 5(k), it can be seen that the triangles outside the wall contour are all unstable obtuse triangle features, so all unstable obtuse triangles in the figure are removed. Figure 5(d) 、 5(h) As shown in 5(i), the remaining acute triangles are all located on the inner side of the tunnel wall. The local minimum width d of the tunnel is estimated by calculating the height of the shortest side of each acute triangle. w .

[0079] The length of each acute triangle is equal to the local minimum width d of the roadway. w Compare, if any side of the triangle is less than d w , then remove the triangle. After removing all triangles with side lengths less than d w After the triangle, Figure 5(d) 、 5(h) As shown in Figure 5(l), the long corridor has no three sides larger than the local minimum width d of the lane. w The T-intersection has only one side whose three sides are greater than the local minimum width d of the roadway. w The triangle at the intersection has only two sides whose three sides are greater than the local minimum width d of the roadway. w triangle.

[0080] As shown in Figure 5(h), the geometric invariant point of the T-intersection is located at the center of the circumscribed circle of the acute triangle, that is, the point shown. As shown in Figure 5(l), the midpoint of the line connecting the centers of the circumscribed circles of the two acute triangles of the cross-shaped intersection is considered to be the location of the geometric invariant point of the cross-shaped intersection. Similarly, for the intersection containing N w The star-shaped intersection of the walls is eliminated by removing obtuse triangles and triangles whose shortest side is less than the local minimum width d of the lane. w After the acute triangle, only N w -2 stable acute triangles remain. The common feature of these stable acute triangles is that their three vertices are located on different walls. w -The geometric mean of the centers of the circumscribed circles of the two acute triangles is used as the geometric fixed point of the star intersection.

[0081] An embodiment of the present invention provides a method and system for extracting geometric fixed points at intersections in underground tunnels, which overcomes the problem that during the flight of an unmanned aerial vehicle equipped with a laser radar, the point cloud structural features of the same intersection scanned by the laser radar undergo huge changes due to large changes in posture.

[0082] Those skilled in the art will appreciate that, in addition to implementing the system and its various devices, modules, and units provided by the present invention in purely computer-readable program code, it is entirely possible to implement the same functions of the system and its various devices, modules, and units provided by the present invention in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers by logically programming the method steps. Therefore, the system and its various devices, modules, and units provided by the present invention can be considered a hardware component, and the devices, modules, and units included therein for implementing various functions can also be considered as structures within the hardware component; the devices, modules, and units for implementing various functions can also be considered as both software modules implementing the method and structures within the hardware component.

[0083] The above describes specific embodiments of the present invention. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art may make various changes or modifications within the scope of the claims, which do not affect the essence of the present invention. The embodiments of this application and the features in the embodiments may be combined with each other in any manner unless there is a conflict.

Claims

1. A method for extracting geometric fixed points at intersections of underground tunnels, characterized in that: include: Step S1: Input a 3D point cloud of the intersection to be processed, and obtain a fitted ground plane equation and mine tunnel height information based on the RANSAC plane fitting algorithm; based on the ground plane equation and the mine tunnel height information, segment the 3D point cloud of the intersection to be processed into ground points and non-ground points, and extract the most structurally distinct and representative wall points from the non-ground points; Step S2: Projecting the wall points from the laser radar coordinate system to the ground plane coordinate system to obtain a wall point cloud after coordinate transformation; Step S3: converting the wall point cloud after coordinate conversion into a two-dimensional occupancy grid image, and using the Delaunay triangulation method to extract the geometric fixed points of the intersection; The step S3 comprises: Step S3.1: Project the wall point cloud onto the ground plane G P w Dimensionality reduction to X g O G Y G plane, converted to a binary image with a resolution of δ×δ, and retaining k×km 2 The square area is taken as the region of interest, where the origin O G is the center of the square area, and the square area is discretized into A binary image of Step S3.2: Use binary unification filtering to remove the smaller isolated areas in the binary image and extract the pixel positions of the remaining continuous areas to form a two-dimensional point set P 2d ; Step S3.3: For the two-dimensional point set P 2d Perform Delaunay triangulation and calculate the local minimum width d of the lane based on the triangulation results w ; The step S3.3 includes: The triangles outside the wall outline are all unstable obtuse triangle features. All unstable obtuse triangles are removed; the remaining acute triangles are all located on the inside of the tunnel wall. The local minimum width d of the tunnel is estimated by calculating the height of the shortest side of each acute triangle. w ; The length of each acute triangle is equal to the local minimum width d of the roadway. w Compare, if any side of the triangle is less than d w , then remove the triangle; After removing all edges less than d w After the triangle is formed, the long corridor has no three sides greater than the local minimum width d of the roadway. w The triangle is retained, and the T-intersection has only one triangle with three sides greater than the local minimum width d of the roadway. w The triangle at the intersection has only two sides whose three sides are greater than the local minimum width d of the roadway. w The triangle is retained for the calculation of invariant points; The geometrically invariant point of a T-shaped intersection is located at the center of the circumscribed circle of the acute triangle. The midpoint of the line connecting the centers of the circumscribed circles of the two triangles of a cross-shaped intersection is considered to be the location of the geometrically invariant point of the cross-shaped intersection. Similarly, for intersections containing N w The star-shaped intersection of the walls is eliminated by removing obtuse triangles and triangles whose shortest side is less than the local minimum width d of the lane. w After the acute triangle, only N w -2 stable acute triangles remain. The common feature of these stable acute triangles is that their three vertices are located on different walls. w -The geometric mean of the centers of the circumscribed circles of the two acute triangles is used as the geometric fixed point of the star intersection.

2. The method for extracting geometric fixed points at underground tunnel intersections according to claim 1, characterized in that: The step S1 comprises: Input the 3D point cloud of the intersection to be processed L P, the three-dimensional point cloud L The points in P are sorted by z value, and Nseed points with the smallest z value are selected as seed points to estimate the initial plane equation G0 of the ground; Set the plane fitting distance threshold ε p , calculate the 3D point cloud L The calculated distance from all points in P to the initial plane equation G0 is selected to be less than the threshold ε p The point is used as a new seed point to update the initial plane equation G0; repeat N iterations g After times, the final fitting ground plane G is obtained plane ; According to the fitted ground plane equation G plane , original 3D point cloud L P is segmented into non-ground points L P ng and ground points L P g ; Calculate non-ground points L P ng Distance d from the ground g , minus the ground distance d g >Height information of mine tunnel h mime Points, thereby segmenting the wall points L P w .

3. A geometric fixed point extraction system for underground tunnel intersections, characterized by: include: Module M1: Inputs the 3D point cloud of the intersection to be processed, obtains the fitted ground plane equation and mine tunnel height information based on the RANSAC plane fitting algorithm; based on the ground plane equation and mine tunnel height information, segments the 3D point cloud of the intersection to be processed into ground points and non-ground points, and extracts the most structurally distinct and representative wall points from the non-ground points; Module M2: Projecting the wall points from the laser radar coordinate system to the ground plane coordinate system to obtain a wall point cloud after coordinate transformation; Module M3: Convert the wall point cloud after coordinate conversion into a two-dimensional occupancy grid image and use the Delaunay triangulation method to extract the geometric fixed points of the intersection; The module M3 includes: Module M3.1: Wall point cloud projected onto the ground plane G P w Dimensionality reduction to X G O G Y G plane, converted to a binary image with a resolution of δ×δ, and retaining k×km 2 The square area is taken as the region of interest, where the origin O G is the center of the square area, and the square area is discretized into A binary image of Module M3.2: Use binary unification filtering to remove smaller isolated areas in the binary image and extract the pixel positions of the remaining continuous areas to form a two-dimensional point set P 2d ; Module M3.3: For a two-dimensional point set P 2d Perform Delaunay triangulation and calculate the local minimum width d of the lane based on the triangulation results w ; The module M3.3 includes: The triangles outside the wall outline are all unstable obtuse triangle features. All unstable obtuse triangles are removed; the remaining acute triangles are all located on the inside of the tunnel wall. The local minimum width d of the tunnel is estimated by calculating the height of the shortest side of each acute triangle. w ; The length of each acute triangle is equal to the local minimum width d of the roadway. w Compare, if any side of the triangle is less than d w , then remove the triangle; After removing all edges less than d w After the triangle is formed, the long corridor has no three sides greater than the local minimum width d of the roadway. w The triangle is retained, and the T-intersection has only one triangle with three sides greater than the local minimum width d of the roadway. w The triangle at the intersection has only two sides whose three sides are greater than the local minimum width d of the roadway. w The triangle is retained for the calculation of invariant points; The geometrically invariant point of a T-shaped intersection is located at the center of the circumscribed circle of the acute triangle. The midpoint of the line connecting the centers of the circumscribed circles of the two triangles of a cross-shaped intersection is considered to be the location of the geometrically invariant point of the cross-shaped intersection. Similarly, for intersections containing N w The star-shaped intersection of the walls is eliminated by removing obtuse triangles and triangles whose shortest side is less than the local minimum width d of the lane. w After the acute triangle, only N w -2 stable acute triangles remain. The common feature of these stable acute triangles is that their three vertices are located on different walls. w -The geometric mean of the centers of the circumscribed circles of the two acute triangles is used as the geometric fixed point of the star intersection.

4. The underground tunnel intersection geometric fixed point extraction system according to claim 3 is characterized in that: The module M1 includes: Input the 3D point cloud of the intersection to be processed L P, the three-dimensional point cloud L The points in P are sorted by z value, and Nseed points with the smallest z value are selected as seed points to estimate the initial plane equation G0 of the ground; Set the plane fitting distance threshold ε p , calculate the 3D point cloud L The calculated distance from all points in P to the initial plane equation G0 is selected to be less than the threshold ε p The point is used as a new seed point to update the initial plane equation G0; repeat N iterations g After times, the final fitting ground plane G is obtained plane ; According to the fitted ground plane equation G plane , original 3D point cloud L P is segmented into non-ground points L P ng and ground points L P g ; Calculate non-ground points L P ng Distance d from the ground g , minus the ground distance d g >Height information of mine tunnel h mine Points, thereby segmenting the wall points L P w .

5. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method for extracting geometric fixed points at intersections of underground tunnels as described in any one of claims 1 to 2 are implemented.

6. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the computer program is executed by a processor, the steps of the method for extracting geometric fixed points at intersections of underground tunnels as described in any one of claims 1 to 2 are implemented.

Images