Automatic identification method for dangerous and easy-to-slip three-dimensional point cloud body block of rock mass

Abstract

The invention provides an automatic identification method for a dangerous and easily-sliding three-dimensional point cloud body block of a rock mass, and belongs to the technical field of mineral engineering. The method comprises the following steps: firstly, automatically identifying rock mass surface structural surfaces by utilizing a three-dimensional Gaussian mixture model algorithm, obtaining each classification cluster in each group of rock mass surface structural surfaces, and storing each classification cluster in all rock mass surface structural surfaces in a specified index mode; secondly, searching adjacent clusters of each classification cluster in all rock mass surface structural surfaces, determining vertexes of rock mass candidate body blocks, and automatically identifying the rock mass candidate body blocks by utilizing a body block growth algorithm; and finally, judging a spatial topological relation between the rock mass candidate body block and the surrounding rock mass candidate body block or the rock mass surface structure so as to judge a dangerous body block, namely the dangerous three-dimensional point cloud body block easy to slip off. The method provided by the invention can be comprehensively suitable for identifying the dangerous and easily-sliding three-dimensional point cloud body block of the rock mass slope in a complex scene such as a mine; the method is easy to implement and remarkable in effect, and the application requirement is met.

Description

Technical Field

[0001] The invention relates to the technical field of mining engineering, and in particular to a method for automatically identifying three-dimensional point cloud blocks of dangerous and easily sliding rock masses. Background Art

[0002] There are natural fractures, folds, cleavages and other different types of geological structures in rock slopes, such as linear, surface and body shapes. Under the action of external stress and affected by geological structures, it is very easy to cause the rock structure to become unstable, resulting in landslides or collapses, which will seriously damage the natural ecological environment and even threaten the safety of human life and property. Therefore, the dangerous and easy-to-slide three-dimensional blocks in rock slopes are a problem that construction workers and engineers are very concerned about.

[0003] Traditional block identification and block information acquisition are usually contact measurements. Surveyors use handheld tools such as geological compasses, measuring ropes, and rulers to obtain indicators such as the occurrence, spacing, and trace length of the accessible rock surface. However, in actual operation, traditional measurement methods will produce large errors in the results due to the complexity of surveyors and on-site conditions, and are time-consuming, labor-intensive, and dangerous. The scale of modern rock mass engineering is expanding, and the construction speed is getting faster and faster. Traditional measurement methods cannot meet modern measurement needs; in addition, for complex construction sites such as high and steep slopes, mine slopes, and island slopes, it is difficult for surveyors to reach the work area using traditional measurement methods. With the rapid development of measurement technology, non-contact measurement has been favored by more researchers, such as drones, three-dimensional laser scanning, digital cameras, etc., which can quickly obtain rock mass characterization data without contacting the rock surface. Compared with traditional methods, it is more time-saving and labor-saving, and has lower risks. It is suitable for the safe and efficient acquisition of original information on the structural surface of modern large-scale geotechnical engineering.

[0004] So far, researchers have rarely studied block identification, and the methods used mainly include flood filling method, block search method, Alpha shape algorithm, depth-first algorithm, etc. At the same time, some scholars manually delineate rock blocks through various processing software. In existing research, only regular hexahedrons and irregular blocks can be obtained, but not all of these blocks are dangerous and prone to sliding. Most of them do not have the risk of sliding and will not have a direct impact on rock slope landslides. Among them, the dangerous and easy-to-slide blocks that affect the occurrence of landslides account for a small part and are difficult to judge. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide a method for automatically identifying dangerous and prone to sliding three-dimensional point cloud blocks of rock mass in response to the deficiencies of the above-mentioned prior art, which uses a block growth algorithm to automatically identify candidate rock mass blocks and judge and identify dangerous and prone to sliding three-dimensional point cloud blocks.

[0006] In order to solve the above technical problems, the technical solution adopted by the present invention is: a method for automatically identifying dangerous and easy-to-slide three-dimensional point cloud blocks of rock mass, comprising the following steps:

[0007] Step 1: Use the three-dimensional Gaussian mixture model algorithm to automatically identify the rock surface structural planes, and obtain each classification cluster in each group of rock surface structural planes, and store each classification cluster in all rock surface structural planes in a specified indexing manner;

[0008] Step 2: Find the adjacent clusters of each classification cluster in all rock surface structural planes, determine the vertices of the rock candidate blocks, and automatically identify the rock candidate blocks using the block growth algorithm, including the following steps:

[0009] Step 2.1: Find the adjacent clusters of each classification cluster in all rock surface structural planes and store them in the adjacent cluster list of the search cluster;

[0010] Take one of the clusters To start the search cluster, use the neighborhood search method to traverse all the remaining classification clusters and obtain the adjacent clusters of all the remaining classification clusters List of neighboring clusters that coexist with the search cluster Where i is the serial number of the rock surface structural surface, and j is the serial number of the cluster in each group of rock surface structural surfaces;

[0011] Step 2.2: Find three adjacent classification clusters, determine the vertices of the rock mass candidate blocks, and store them in the rock mass candidate block vertex list verticesList;

[0012] Find three adjacent clusters, one of which To start the search cluster, traverse the cluster The list of neighboring clusters of the search cluster Simultaneously search the neighboring cluster list The list of neighboring clusters for each cluster in The intersection points of three adjacent classification clusters are used as the vertices of the rock mass candidate blocks; in the same way, the vertices of all rock mass candidate blocks are obtained and stored in the rock mass candidate block vertex list verticesList;

[0013] Step 2.3: Using the block growth algorithm to identify candidate rock blocks, including the following steps:

[0014] Step 2.3.1: Establish the initial space of the candidate rock mass block, calculate the opening direction of the initial space of the candidate rock mass block, and compare the angle between the opening direction and the normal direction with 90 degrees to determine whether the initial state of the candidate rock mass block meets the conditions; index other vertices on the ridge where the vertices that meet the initial state of the candidate rock mass block are located, and execute step 2.3.2 for the next step of judgment;

[0015] Calculate the inclination, dip angle and normal direction of the original rock mass surface, and define the normal direction V of the original rock mass surface n Towards the inside of the rock mass surface; the rock mass candidate block initially grows from a vertex, and the first vertex is formed by the intersection of three rock mass surfaces. The initial three surfaces can be used to determine a unique space, namely, the initial space of the rock mass candidate block;

[0016] Calculate the opening direction V of the initial space of the rock mass candidate block i ; When V i With V n When the angle is greater than 90 degrees, it means that this vertex is not a valid starting point and does not meet the starting state of the rock mass candidate block, and the next intersection point is determined; when V i With V n When the angle is less than 90 degrees, it means that this vertex is a valid starting point and meets the starting state of the candidate rock mass block;

[0017] Add the vertices that meet the initial state of the rock candidate block to the vertex list verticesList of the rock candidate block, add the first cluster to the plane list planesList of the rock candidate block, and index the other vertices on the edge line where the vertices that meet the initial state of the rock candidate block are located, and execute step 2.3.2 for the next step of judgment; the edge line is the intersection line of two rock surface;

[0018] Step 2.3.2: Determine whether the vertex on the endpoint of the edge is the endpoint of the edge. If the vertex is on the endpoint of the edge, execute step 2.3.3 for the next step of determination.

[0019] Each edge line contains multiple vertices. If the vertex is inside the edge line, determine the next vertex in the rock mass candidate block vertex list verticesList; if the vertex is at the endpoint of the edge line, execute step 2.3.3 to make the next judgment;

[0020] Step 2.3.3: Determine whether the vertex at the edge line endpoint is on the opening side of the rock mass candidate block. If so, add the vertex at the edge line endpoint to the rock mass candidate block vertex list verticesList;

[0021] If the vertex at the edge endpoint is outside the rock candidate block, the vertex at the edge endpoint does not belong to the rock candidate block, and the next vertex is determined; if the vertex at the edge endpoint is inside the rock candidate block, the vertex at the edge endpoint belongs to the candidate block, the vertex at the edge endpoint is added to the rock candidate block vertex list verticesList, and the other vertices on the edge where the vertex is located are indexed, and step 2.3.4 is executed for the next step of determination;

[0022] Step 2.3.4: Determine whether the cluster contained in the newly added vertex already exists in the candidate block. If not, add the new cluster to the rock candidate block surface list planesList;

[0023] Each vertex contains three clusters. It is determined whether the three clusters corresponding to the vertex at the edge endpoint already exist in the rock candidate block list. If not, the new cluster is added to the rock candidate block surface list planesList, and the number of clusters k becomes k+1. Similarly, all clusters are traversed until all candidate blocks are indexed, and then step 2.3.5 is executed to make the next judgment. The rock candidate block list includes the rock candidate block surface list planesList and the rock candidate block vertex list verticesList.

[0024] Step 2.3.5: Determine whether there are duplicate blocks in the candidate rock mass blocks. If so, delete the duplicate blocks. Otherwise, execute step 3.

[0025] At this time, the rock candidate block consists of a vertex list verticesList of the rock candidate block and a surface list planesList of the rock candidate block. The number of vertices and the order of vertices in the vertex list verticesList of the rock candidate block are judged. If the number of vertices is the same but the order is different, the rock candidate block is retained, and the number of rock candidate blocks m is changed to m+1, until all rock candidate blocks are judged;

[0026] Step 3: Determine the spatial topological relationship between the candidate rock mass blocks and the surrounding candidate rock mass blocks or the surface structure of the rock mass, so as to determine the dangerous blocks, that is, the dangerous and easy-to-slide three-dimensional point cloud blocks;

[0027] Index the lower surface of each candidate block; if the lower surface of a candidate block does not intersect with any position of the rock surface, mark this rock candidate block as a dangerous block; and so on, judge all rock candidate blocks, mark all dangerous blocks, and all marked dangerous blocks are called dangerous and easy to slide three-dimensional point cloud blocks.

[0028] The beneficial effects of adopting the above technical scheme are: the present invention provides a method for automatically identifying three-dimensional point cloud blocks of dangerous rock masses that are prone to sliding, which can automatically identify three-dimensional point cloud blocks of dangerous rock masses that are prone to sliding, and can be fully applicable to the identification of three-dimensional point cloud blocks of dangerous rock slopes that are prone to sliding in complex scenes such as mines; the method of the present invention is simple to implement, has significant effects, and meets the application requirements. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 A flow chart of a method for automatically identifying dangerous and prone-to-slide three-dimensional point cloud blocks of rock masses provided by an embodiment of the present invention;

[0030] Figure 2 A flow chart of a block growth algorithm provided by an embodiment of the present invention;

[0031] Figure 3 A process diagram of identifying candidate rock blocks using a block growth algorithm provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0032] The specific implementation of the present invention is further described in detail below in conjunction with the accompanying drawings and examples. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0033] In this embodiment, a method for automatically identifying dangerous and prone to sliding three-dimensional point cloud blocks of rock mass is provided. Figure 1 As shown, the following steps are included:

[0034] Step 1: Use the three-dimensional Gaussian mixture model algorithm to automatically identify the rock surface structural planes, and obtain each classification cluster in each group of rock surface structural planes, and store each classification cluster in all rock surface structural planes in a specified indexing manner;

[0035] Step 2: Find the adjacent clusters of each classification cluster in all rock surface structural planes, determine the vertices of the rock candidate blocks, and automatically identify the rock candidate blocks using the block growth algorithm, including the following steps:

[0036] Step 2.1: Find the adjacent clusters of each classification cluster in all rock surface structural planes and store them in the adjacent cluster list of the search cluster;

[0037] Take one of the clusters To start the search cluster, use the neighborhood search method to traverse all the remaining classification clusters and obtain the adjacent clusters of all the remaining classification clusters List of neighboring clusters that coexist with the search cluster Where i is the serial number of the rock surface structural surface, and j is the serial number of the cluster in each group of rock surface structural surfaces;

[0038] Step 2.2: Find three adjacent classification clusters, determine the vertices of the rock mass candidate blocks, and store them in the rock mass candidate block vertex list verticesList;

[0039] Find three adjacent clusters, one of which To start the search cluster, traverse the cluster The list of neighboring clusters of the search cluster Simultaneously search the neighboring cluster list The list of neighboring clusters for each cluster in If the adjacent cluster list and a list of adjacent clusters contains the same adjacent clusters, such as Then explain For three adjacent clusters, the three adjacent classification clusters are intersected to obtain the intersection point, and the intersection point is the vertex of the rock mass candidate block; in the same way, the vertices of all rock mass candidate blocks are obtained and stored in the rock mass candidate block vertex list verticesList;

[0040] Step 2.3: Use the block growth algorithm (BGA) to identify candidate rock blocks, such as Figure 2-3 As shown, the following steps are included:

[0041] Step 2.3.1: Establish the initial space of the candidate rock mass block, calculate the opening direction of the initial space of the candidate rock mass block, and compare the angle between the opening direction and the normal direction with 90 degrees to determine whether the initial state of the candidate rock mass block meets the conditions; index other vertices on the ridge where the vertices that meet the initial state of the candidate rock mass block are located, and execute step 2.3.2 for the next step of judgment;

[0042] Calculate the inclination, dip angle and normal direction of the original rock mass surface, and define the normal direction V of the original rock mass surface n Towards the inside of the rock mass surface; the rock mass candidate block initially grows from a vertex, and the first vertex is formed by the intersection of three surfaces. The initial three surfaces can be used to determine a unique space, that is, the initial space of the rock mass candidate block;

[0043] Calculate the opening direction V of the initial space of the rock mass candidate block i ; When V i With V n When the angle is greater than 90 degrees, it means that this vertex is not a valid starting point and does not meet the starting state of the rock mass candidate block, and the next intersection point is determined; when V i With V n When the angle is less than 90 degrees, it means that this vertex is a valid starting point and meets the starting state of the candidate rock mass block;

[0044] Add the vertices that meet the initial state of the rock candidate block to the vertex list verticesList of the rock candidate block, add the first cluster to the plane list planesList of the rock candidate block, and index the other vertices on the edge line where the vertices that meet the initial state of the rock candidate block are located, and execute step 2.3.2 for the next step of judgment; the edge line is the intersection line of two rock surface;

[0045] Step 2.3.2: Determine whether the vertex on the endpoint of the edge is the endpoint of the edge. If the vertex is on the endpoint of the edge, execute step 2.3.3 for the next step of determination.

[0046] Each edge line contains multiple vertices. If the vertex is inside the edge line, determine the next vertex in the rock mass candidate block vertex list verticesList; if the vertex is at the endpoint of the edge line, execute step 2.3.3 to make the next judgment;

[0047] Step 2.3.3: Determine whether the vertex at the edge line endpoint is on the opening side of the rock mass candidate block. If so, add the vertex at the edge line endpoint to the rock mass candidate block vertex list verticesList;

[0048] If the vertex at the edge endpoint is outside the rock candidate block, the vertex at the edge endpoint does not belong to the rock candidate block, and the next vertex is determined; if the vertex at the edge endpoint is inside the rock candidate block, the vertex at the edge endpoint belongs to the candidate block, the vertex at the edge endpoint is added to the rock candidate block vertex list verticesList, and the other vertices on the edge where the vertex is located are indexed, and step 2.3.4 is executed for the next step of determination;

[0049] Step 2.3.4: Determine whether the cluster contained in the newly added vertex already exists in the candidate block. If not, add the new cluster to the rock candidate block surface list planesList;

[0050] Each vertex contains three clusters. It is determined whether the three clusters corresponding to the vertices at the edge endpoints already exist in the rock candidate block list. If not, the new cluster is added to the rock candidate block surface list planesList, and the number of clusters k becomes k+1. Similarly, all clusters are traversed until all candidate blocks are indexed, and then step 2.3.5 is executed to make the next judgment. The rock candidate block list includes the rock candidate block surface list planesList and the rock candidate block vertex list verticesList.

[0051] Step 2.3.5: Determine whether there are duplicate blocks in the candidate rock mass blocks. If so, delete the duplicate blocks. Otherwise, execute step 3.

[0052] At this time, the rock candidate block consists of a vertex list verticesList of the rock candidate block and a surface list planesList of the rock candidate block. The number of vertices and the order of vertices in the vertex list verticesList of the rock candidate block are judged. If the number of vertices is the same but the order is different, the rock candidate block is retained, and the number of rock candidate blocks m is changed to m+1, until all rock candidate blocks are judged;

[0053] Step 3: Determine the spatial topological relationship between the candidate rock mass blocks and the surrounding candidate rock mass blocks or the surface structure of the rock mass, so as to determine the dangerous blocks, that is, the dangerous and easy-to-slide three-dimensional point cloud blocks;

[0054] Index the lower surface of each candidate block; if the lower surface does not intersect with any position of the rock surface, mark this rock candidate block as a dangerous block; and so on, judge all rock candidate blocks, mark all dangerous blocks, and all marked dangerous blocks are called dangerous and easy to slide three-dimensional point cloud blocks.

[0055] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some or all of the technical features therein. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope defined by the claims of the present invention.

Claims

1. A method for automatically identifying dangerous and prone-to-slide three-dimensional point cloud blocks of rock mass, characterized by: The following steps are involved: Step 1: Use the three-dimensional Gaussian mixture model algorithm to automatically identify the rock surface structural planes, and obtain each classification cluster in each group of rock surface structural planes, and store each classification cluster in all rock surface structural planes in a specified indexing manner; Step 2: Find the adjacent clusters of each classification cluster in all rock surface structural planes, determine the vertices of the rock candidate blocks, and automatically identify the rock candidate blocks using the block growth algorithm; Step 3: Determine the spatial topological relationship between the candidate rock mass blocks and the surrounding candidate rock mass blocks or the surface structure of the rock mass, so as to determine the dangerous blocks, that is, the dangerous and easy-to-slide three-dimensional point cloud blocks.

2. The method for automatically identifying dangerous and prone-to-slide three-dimensional point cloud blocks of rock mass according to claim 1, characterized in that: The step 2 comprises the following steps: Step 2.1: Find the adjacent clusters of each classification cluster in all rock surface structural planes and store them in the adjacent cluster list of the search cluster; Step 2.2: Find three adjacent classification clusters, determine the vertices of the rock mass candidate blocks, and store them in the rock mass candidate block vertex list verticesList; Step 2.3: Use the block growing algorithm to identify candidate rock blocks.

3. The method for automatically identifying dangerous and prone-to-slide three-dimensional point cloud blocks of rock mass according to claim 2, characterized in that: The specific method of step 2.1 is: Take one of the clusters To start the search cluster, use the neighborhood search method to traverse all the remaining classification clusters and obtain the adjacent clusters of all the remaining classification clusters List of neighboring clusters that coexist with the search cluster Where i is the serial number of the rock surface structural surface, and j is the serial number of the cluster in each group of rock surface structural surfaces.

4. The method for automatically identifying dangerous and prone-to-slide three-dimensional point cloud blocks of rock mass according to claim 3 is characterized by: The specific method of step 2.2 is: Find three adjacent clusters, one of which To start the search cluster, traverse the cluster The list of neighboring clusters of the search cluster Simultaneously search the neighboring cluster list The list of neighboring clusters for each cluster in The intersection points of three adjacent classification clusters are taken as the vertices of the candidate rock mass blocks; In the same way, the vertices of all rock mass candidate blocks are obtained and stored in the rock mass candidate block vertex list verticesList.

5. The method for automatically identifying dangerous and prone-to-slide three-dimensional point cloud blocks of rock mass according to claim 4, characterized in that: The step 2.3 comprises the following steps: Step 2.3.1: Establish the initial space of the candidate rock mass block, calculate the opening direction of the initial space of the candidate rock mass block, and compare the angle between the opening direction and the normal direction with 90 degrees to determine whether the initial state of the candidate rock mass block meets the conditions; index other vertices on the ridge where the vertices that meet the initial state of the candidate rock mass block are located, and execute step 2.3.2 for the next step of judgment; Step 2.3.2: Determine whether the vertex on the endpoint of the edge is the endpoint of the edge. If the vertex is on the endpoint of the edge, execute step 2.3.3 for the next step of determination; Each edge line contains multiple vertices. If the vertex is inside the edge line, determine the next vertex in the rock mass candidate block vertex list verticesList; if the vertex is at the endpoint of the edge line, execute step 2.3.3 to make the next judgment; Step 2.3.3: Determine whether the vertex at the end point of the edge line is on the opening side of the rock mass candidate block. If so, add the vertex at the end point of the edge line to the rock mass candidate block vertex list verticesList; Step 2.3.4: Determine whether the cluster contained in the newly added vertex already exists in the candidate block. If not, add the new cluster to the rock candidate block surface list planesList; Step 2.3.5: Determine whether there are duplicate blocks in the candidate rock mass blocks. If so, delete the duplicate blocks. Otherwise, execute step 3.

6. The method for automatically identifying dangerous and prone-to-slide three-dimensional point cloud blocks of rock mass according to claim 5, characterized in that: The specific method of step 2.3.1 is: Calculate the inclination, dip angle and normal direction of the original rock mass surface, and define the normal direction V of the original rock mass surface n Towards the inside of the rock mass surface; the rock mass candidate block initially grows from a vertex, and the first vertex is formed by the intersection of three rock mass surfaces. The initial three surfaces can be used to determine a unique space, namely, the initial space of the rock mass candidate block; Calculate the opening direction V of the initial space of the rock mass candidate block i ; When V i With V n When the angle is greater than 90 degrees, it means that this vertex is not a valid starting point and does not meet the starting state of the rock mass candidate block, and the next intersection point is determined; When V i With V n When the angle is less than 90 degrees, it means that this vertex is a valid starting point and meets the starting state of the candidate rock mass block; Add the vertices that meet the starting state of the rock candidate block to the vertex list verticesList of the rock candidate block, add the first cluster to the surface list planesList of the rock candidate block, and index the other vertices on the edge line where the vertices that meet the starting state of the rock candidate block are located, and execute step 2.3.2 for the next judgment; the edge line is the intersection line of two rock surfaces.

7. The method for automatically identifying dangerous and prone-to-slide three-dimensional point cloud blocks of rock mass according to claim 6, characterized in that: The specific method of step 2.3.3 is: If the vertex at the end point of the edge line is outside the rock mass candidate block, the vertex at the end point of the edge line does not belong to the rock mass candidate block, and the next vertex is determined; If the vertex at the edge endpoint is inside the rock candidate block, the vertex at the edge endpoint belongs to the candidate block. The vertex at the edge endpoint is added to the rock candidate block vertex list verticesList, and other vertices on the edge where the vertex is located are indexed. Execute step 2.3.4 for the next judgment.

8. The method for automatically identifying dangerous and prone-to-slide three-dimensional point cloud blocks of rock mass according to claim 7, characterized in that: The specific method of step 2.3.4 is: Each vertex contains three clusters. It is determined whether the three clusters corresponding to the vertex at the edge endpoint already exist in the list of rock candidate blocks. If not, the new cluster is added to the list of rock candidate block surfaces planesList, and the number of clusters k becomes k+1. And so on, traverse all clusters until all candidate blocks are indexed, then execute step 2.3.5 to make the next judgment; the rock candidate block list includes a rock candidate block surface list planesList and a rock candidate block vertex list verticesList.

9. The method for automatically identifying dangerous and prone-to-slide three-dimensional point cloud blocks of rock mass according to claim 8, characterized in that: The specific method of step 2.3.5 is: At this time, the rock candidate blocks are composed of the rock candidate block vertex list verticesList and the rock candidate block surface list planesList. The number of vertices and the order of vertices in the rock candidate block vertex list verticesList are judged. When the number of vertices is the same but the order is different, the rock candidate blocks are retained, and the number of rock candidate blocks m becomes m+1, until all rock candidate blocks are judged.

10. The method for automatically identifying dangerous and prone-to-slide three-dimensional point cloud blocks of rock mass according to claim 9, characterized in that: The specific method of step 3 is: Index the lower surface of each candidate block; if the lower surface of a candidate block does not intersect with any position of the rock surface, mark this rock candidate block as a dangerous block; and so on, judge all rock candidate blocks, mark all dangerous blocks, and all marked dangerous blocks are called dangerous and easy to slide three-dimensional point cloud blocks.

Images