A method, device and computer program product for detecting under-excavation in high-speed railway tunnel construction

Abstract

The present invention belongs to the technical field of tunnel construction detection. To solve the problems of large workload and low data acquisition efficiency existing in the prior art, a method, device and computer program product for detecting under-excavation in high-speed railway tunnel construction are disclosed. The present invention uses three-dimensional point cloud data to detect tunnel under-excavation. Its acquisition speed is fast and can be directly scanned by a laser scanner without point-by-point measurement. Therefore, not only the workload is reduced, but also the detection efficiency is improved. At the same time, the three-dimensional point cloud data can reflect the geometric state of the tunnel and can construct an accurate construction excavation section during detection. In this way, the accuracy of detection can also be guaranteed. Based on this, the present invention provides a tunnel under-excavation detection method with high efficiency, low workload and high precision, which is very suitable for large-scale application and promotion.

Description

Technical Field

[0001] The present invention belongs to the technical field of tunnel construction detection, and particularly relates to a method, device and computer program product for detecting under-excavation in high-speed railway tunnel construction. Background Art

[0002] In tunnel construction by drilling and blasting method, over-excavation and under-excavation of the tunnel are extremely serious safety hazards, which seriously affect the quality of tunnel excavation and subsequent construction. Among them, over-excavation and under-excavation of the tunnel are judged based on the designed excavation contour line. That is, the part of the actually excavated cross-section outside the reference line is over-excavation, and the part inside the reference line is called under-excavation. Over-excavation will cause waste of materials in the advanced support during the later backfilling process, and at the same time, the support effect is not good, there are safety hazards. The consequence of under-excavation is that it may cause greater trouble in transportation and support, and sometimes it is even necessary to carry out secondary blasting for correction, which also wastes materials and manpower. Therefore, reducing the occurrence of over-excavation and under-excavation, timely detecting the situation of over-excavation and under-excavation and taking corresponding remedial measures have great engineering significance.

[0003] At present, tunnel under-excavation detection mainly uses equipment such as section meters and total stations to measure the excavation cross-section point by point and surface by surface. Although this traditional method can meet the requirements of under-excavation detection, its "point-by-point" measurement method (that is, manually operating the instrument to measure point by point or line by line) not only has a large workload, but also results in low data acquisition efficiency, thus affecting the construction progress. Therefore, based on the above deficiencies, how to provide a tunnel under-excavation detection method with high efficiency and low workload has become an urgent problem to be solved. Summary of the Invention

[0004] The purpose of the present invention is to provide a method, device and computer program product for detecting under-excavation in high-speed railway tunnel construction, so as to solve the problems of large workload and low data acquisition efficiency existing in the prior art.

[0005] To achieve the above purpose, the present invention adopts the following technical solutions:

[0006] In the first aspect, a method for detecting under-excavation in high-speed railway tunnel construction is provided, including:

[0007] Obtaining a three-dimensional point cloud data set of the target construction tunnel and a three-dimensional surface model of the theoretical excavation contour of the target construction tunnel;

[0008] Based on the three-dimensional point cloud data set, constructing a three-dimensional surface model of the actual excavation contour of the target construction tunnel, wherein both the three-dimensional surface model of the actual excavation contour and the three-dimensional surface model of the theoretical excavation contour are triangular mesh models;

[0009] For any vertex in the three-dimensional surface model of the actual excavation profile, determine the triangular patch closest to the vertex from the three-dimensional surface model of the theoretical excavation profile, and use it as the under-excavation detection reference plane corresponding to the vertex. Then, take the distance between the vertex and the under-excavation detection reference plane as the construction error distance;

[0010] Determine whether the construction error distance is 0;

[0011] If not, determine the sign of the construction error distance according to the vector of the normal line of the vertex to the under-excavation detection reference plane;

[0012] If the sign of the construction error distance is negative, determine that the vertex is an under-excavated point. After polling all the vertices in the three-dimensional surface model of the actual excavation profile, obtain the under-excavation detection result of the target construction tunnel.

[0013] Based on the above disclosure, the present invention first obtains the three-dimensional point cloud dataset of the target construction tunnel and the three-dimensional surface model of the theoretical excavation profile of the target construction tunnel. Then, based on the aforementioned three-dimensional point cloud dataset, construct the three-dimensional surface model of the actual excavation profile of the target construction tunnel, and both of the aforementioned models are triangular mesh models. Based on this, the under-excavation detection of the tunnel can be carried out according to the three-dimensional surface model of the actual excavation profile and the three-dimensional surface model of the theoretical excavation profile, that is, take the three-dimensional surface model of the theoretical excavation profile as the reference model and the three-dimensional surface model of the actual excavation profile as the comparison model. Then, convert the under-excavation detection into calculating the distance between the comparison model and the reference model, that is, by determining the triangular patch closest to each vertex in the comparison model in the reference model, and then taking the closest triangular patch as the under-excavation detection reference plane (i.e., the reference plane) corresponding to each vertex. Next, determine whether the position of each vertex in the comparison model is an under-excavated point by determining the positive and negative of the distance between each vertex and the corresponding reference plane.

[0014] Specifically, the present invention determines the sign of the distance from each vertex to the corresponding reference plane by means of the vector between the normal lines of each vertex to the corresponding reference plane. Among them, if the distance from any vertex to the corresponding reference plane is negative, it means that the actual excavation point corresponding to the vertex has not reached the position of the theoretical excavation point, and the position of the vertex is an under-excavated position. On the contrary, if the distance is positive, it means it is an over-excavated position, and if the distance is 0, it means that the excavation position of the vertex is on the reference plane. Thus, through the above method, the under-excavation detection result of the target construction tunnel can be obtained quickly.

[0015] Through the above design, the present invention constructs a three-dimensional surface model of the actual excavation contour of the tunnel by collecting the three-dimensional point cloud dataset of the target construction tunnel. Then, from the three-dimensional surface model of the theoretical excavation contour of the tunnel, the triangular patches that are closest to each vertex in the actual model are determined and used as the under-excavation detection reference planes for each vertex. Finally, by calculating the positive or negative of the distance from the vertex to the corresponding reference plane, it is determined whether each vertex in the actual model is an under-excavation point, thereby obtaining the final under-excavation detection result. Thus, using three-dimensional point cloud data for tunnel under-excavation detection has a fast acquisition speed and can be directly scanned by a laser scanner without the need for point-by-point measurement. Therefore, it not only reduces the workload but also improves the detection efficiency. At the same time, the three-dimensional point cloud data can reflect the geometric state of the tunnel and can construct an accurate construction excavation section during detection. In this way, the accuracy of the detection can also be ensured. Based on this, the present invention provides a tunnel under-excavation detection method with high efficiency, low workload, and high precision, which is very suitable for large-scale application and promotion.

[0016] In a possible design, constructing the three-dimensional surface model of the actual excavation contour of the target construction tunnel based on the three-dimensional point cloud dataset includes:

[0017] Extracting a plurality of feature point clouds from the three-dimensional point cloud dataset to form a feature point cloud set;

[0018] Calculating the local density of each feature point cloud in the feature point cloud set;

[0019] Using the feature point cloud set to construct an initial three-dimensional surface model of the excavation contour of the target construction tunnel;

[0020] According to the local density of each feature point cloud in the initial three-dimensional surface model of the excavation contour, performing triangular patch optimization processing on the initial three-dimensional surface model of the excavation contour to obtain the actual three-dimensional surface model of the excavation contour after the triangular patch optimization processing.

[0021] In a possible design, extracting a plurality of feature point clouds from the three-dimensional point cloud dataset to form a feature point cloud set includes:

[0022] Initializing the sampling iteration number s and obtaining the side length of the voxel grid at the s-th sampling, where when s is 1, the side length of the voxel grid at the s-th time is the initial side length;

[0023] Dividing the three-dimensional point cloud dataset into a number of voxel grids according to the side length of the voxel grid at the s-th sampling, where each voxel grid contains a plurality of three-dimensional point clouds;

[0024] Extracting the feature point cloud at the s-th sampling from the plurality of three-dimensional point clouds in each voxel grid;

[0025] Determine whether the sampling iteration stop condition is satisfied, where the sampling iteration stop condition is that the number of the feature point clouds at the s-th sampling is equal to the preset number;

[0026] If not, update the side length of the voxel grid at the s-th sampling according to the number of the feature point clouds at the s-th sampling and the preset number to obtain the updated side length;

[0027] Increment s by 1, replace the side length of the voxel grid at the s-th sampling with the updated side length, and re-partition the three-dimensional point cloud dataset into a plurality of voxel grids according to the side length of the voxel grid at the s-th sampling until the sampling iteration stop condition is satisfied, so as to form an initial feature point cloud set by using the feature point clouds sampled when the sampling iteration stop condition is satisfied;

[0028] Perform denoising processing on the initial feature point cloud set, so as to obtain the feature point cloud set after the denoising processing.

[0029] In a possible design, extracting the feature point cloud at the s-th sampling from the multiple three-dimensional point clouds in each voxel grid includes:

[0030] For any three-dimensional point cloud in any voxel grid, determine the neighborhood point cloud of the any three-dimensional point cloud from the any voxel grid and the target grid, where the target grid is the voxel grid adjacent to the any voxel grid among the plurality of voxel grids;

[0031] Determine the first normal vector of the any three-dimensional point cloud and the second normal vectors of each neighborhood point cloud;

[0032] Calculate the angles between the first normal vector and each second normal vector to obtain a plurality of normal vector angles;

[0033] Obtain the average value of the plurality of normal vector angles;

[0034] Judge whether the average value is greater than or equal to the angle threshold;

[0035] If so, use the any three-dimensional point cloud as a feature point cloud in the any voxel grid, and after polling all the three-dimensional point clouds in all the voxel grids, obtain the feature point cloud at the s-th sampling.

[0036] In a possible design, updating the side length of the voxel grid at the s-th sampling according to the number of the feature point clouds at the s-th sampling and the preset number to obtain the updated side length includes:

[0037] Calculate the voxel grid side length adjustment coefficient at the s-th sampling according to the number of feature point clouds at the s-th sampling and the preset number;

[0038] Calculate the updated side length according to the voxel grid side length adjustment coefficient and the initial side length at the s-th sampling.

[0039] In a possible design, use the set of feature point clouds to construct the initial excavation contour three-dimensional surface model of the target construction tunnel, including:

[0040] Randomly select three feature point clouds from the set of feature point clouds to form an initial triangle;

[0041] Take any one of the sides of the initial triangle as the growth side, and screen out the optimal extended point cloud corresponding to the growth side from the set of feature point clouds;

[0042] Use the optimal extended point cloud and the growth side to construct an extended triangle;

[0043] Update the initial triangle to the extended triangle, and re-take any one of the sides of the initial triangle as the growth side, and screen out the optimal extended point cloud corresponding to the growth side from the set of feature point clouds until all the feature point clouds in the set of feature point clouds are divided, so as to use the initial triangle and the obtained extended triangles to construct the initial excavation contour three-dimensional surface model.

[0044] In a possible design, screening out the optimal extended point cloud corresponding to the growth side from the set of feature point clouds includes:

[0045] Screen out the feature point cloud closest to the growth side from the set of feature point clouds as the initial extended point cloud;

[0046] Construct an initial extended triangle based on the initial extended point cloud and the growth side;

[0047] Judge whether there is a specified feature point cloud inside the circumcircle of the initial extended triangle, where the specified feature point cloud is all the feature point clouds in the set of feature point clouds except the feature point clouds corresponding to the three vertices of the initial extended triangle;

[0048] If so, randomly select one from the specified feature point clouds as the initial extended point cloud, and re-construct the initial extended triangle based on the initial extended point cloud and the growth side until there is no specified feature point cloud inside the circumcircle of the constructed initial extended triangle, so as to use the initial extended point cloud corresponding to the initial extended triangle without the specified feature point cloud inside the circumcircle as the optimal extended point cloud.

[0049] In a possible design, calculating the local density of each feature point cloud in the feature point cloud set includes:

[0050] For any feature point cloud in the feature point cloud set, obtaining the k-neighborhood point cloud of the any feature point cloud;

[0051] Calculating the distances between the any feature point cloud and each k-neighborhood point cloud, and taking the average value of all the distances as the local density of the any feature point cloud;

[0052] Correspondingly, performing triangular facet optimization processing on the initial excavation contour three-dimensional surface model according to the local densities of the feature point clouds in the initial excavation contour three-dimensional surface model includes:

[0053] For any initial triangular facet in the initial excavation contour three-dimensional surface model, calculating the circumradius of the any initial triangular facet;

[0054] Calculating the density threshold corresponding to the any initial triangular facet according to the local densities of the feature point clouds corresponding to the three endpoints of the any initial triangular facet;

[0055] Judging whether the any initial triangular facet is a segmentation triangular facet according to the circumradius and the density threshold;

[0056] If so, inserting a new feature point cloud into the any initial triangular facet, so as to perform segmentation processing on the any initial triangular facet based on the new feature point cloud, and obtaining the actual excavation contour three-dimensional surface model when all the initial triangular facets in the initial excavation contour three-dimensional surface model are polled.

[0057] In a second aspect, a device for detecting under-excavation in high-speed railway tunnel construction is provided, including:

[0058] An acquisition unit, configured to acquire a three-dimensional point cloud data set of a target construction tunnel and a theoretical excavation contour three-dimensional surface model of the target construction tunnel;

[0059] A model construction unit, configured to construct an actual excavation contour three-dimensional surface model of the target construction tunnel based on the three-dimensional point cloud data set, where both the actual excavation contour three-dimensional surface model and the theoretical excavation contour three-dimensional surface model are triangular mesh models;

[0060] An error detection unit, for any vertex in the three-dimensional surface model of the actual excavation contour, is used to determine, from the three-dimensional surface model of the theoretical excavation contour, the triangular patch closest to the any vertex as the under-excavation detection reference plane for the any vertex, and use the distance between the any vertex and the under-excavation detection reference plane as the construction error distance;

[0061] An under-excavation detection unit is used to determine whether the construction error distance is 0;

[0062] An under-excavation detection unit is used to, when it is determined that the construction error distance is not 0, determine the sign of the construction error distance according to the vector of the normal line of the any vertex to the under-excavation detection reference plane;

[0063] If the sign of the construction error distance is a negative sign, the under-excavation detection unit is further used to determine that the any vertex is an under-excavated point, and after polling all the vertices in the three-dimensional surface model of the actual excavation contour, obtain the under-excavation detection result of the target construction tunnel.

[0064] In a third aspect, another high-speed railway tunnel construction under-excavation detection device is provided. Taking the device as an electronic device as an example, it includes a memory, a processor, and a transceiver that are communicatively connected in sequence. Among them, the memory is used to store a computer program, the transceiver is used to send and receive messages, and the processor is used to read the computer program and execute the high-speed railway tunnel construction under-excavation detection method as described in the first aspect or any possible design in the first aspect.

[0065] In a fourth aspect, a storage medium is provided, on which instructions are stored. When the instructions are run on a computer, the high-speed railway tunnel construction under-excavation detection method as described in the first aspect or any possible design in the first aspect is executed.

[0066] In a fifth aspect, a computer program product containing instructions is provided. When the instructions are run on a computer, the computer is made to execute the high-speed railway tunnel construction under-excavation detection method as described in the first aspect or any possible design in the first aspect.

[0067] Beneficial effects:

[0068] (1) The present invention uses three-dimensional point cloud data for tunnel under-excavation detection. Its acquisition speed is fast and can be directly scanned by a laser scanner without point-by-point measurement. Therefore, not only the workload is reduced, but also the detection efficiency is improved. At the same time, the three-dimensional point cloud data can reflect the geometric state of the tunnel and can construct an accurate construction excavation section during detection. In this way, the accuracy of the detection can also be guaranteed. Based on this, the present invention provides a tunnel under-excavation detection method with fast efficiency, low workload and high precision, and is thus very suitable for large-scale application and promotion. Description of the drawings

[0069] Figure 1 It is a schematic flow chart of the steps of the method for detecting under-excavation in high-speed railway tunnel construction provided by the embodiment of the present invention;

[0070] Figure 2 It is a schematic structural diagram of the device for detecting under-excavation in high-speed railway tunnel construction provided by the embodiment of the present invention;

[0071] Figure 3 It is a schematic structural diagram of the electronic device provided by the embodiment of the present invention. Detailed implementation manners

[0072] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or in the prior art, the following will briefly introduce the present invention in combination with the accompanying drawings and the descriptions of the embodiments or the prior art. Obviously, the following descriptions of the structures of the drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings. It should be noted here that the descriptions of these embodiments are used to help understand the present invention, but do not constitute a limitation to the present invention.

[0073] It should be understood that although terms such as first and second may be used herein to describe various units, these units should not be limited by these terms. These terms are only used to distinguish one unit from another. For example, the first unit may be called the second unit, and similarly, the second unit may be called the first unit, without departing from the scope of the exemplary embodiments of the present invention.

[0074] It should be understood that for the term "and / or" that may appear in this article, it is only a description of the association relationship of the associated objects, indicating that three relationships may exist. For example, A and / or B may represent: A exists alone, B exists alone, and both A and B exist at the same time; for the term " / and" that may appear in this article, it is a description of another association object relationship, indicating that two relationships may exist. For example, A / and B may represent: A exists alone, and both A and B exist; in addition, for the character " / " that may appear in this article, generally, it means that the front and rear associated objects are in an "or" relationship.

[0075] Embodiment:

[0076] See Figure 1As shown in the figure, the overbreak detection method for high-speed railway tunnel construction provided in this embodiment constructs a three-dimensional surface model of the actual excavation contour of the tunnel by collecting the three-dimensional point cloud dataset of the target construction tunnel. Then, from the three-dimensional surface model of the theoretical excavation contour of the tunnel, the triangular patches closest to each vertex in the actual model are determined and used as the overbreak detection reference planes for each vertex. Finally, by calculating the positive or negative of the distance from the vertex to the corresponding reference plane, it is determined whether each vertex in the actual model is an overbreak point, thereby obtaining the final overbreak detection result. In this way, compared with the traditional technology, the three-dimensional point cloud data can be directly obtained by a laser scanner without point-by-point measurement. Therefore, not only the efficiency is improved, but also the measurement workload is reduced. At the same time, a large amount of three-dimensional point cloud data can reflect the actual geometric shape of the tunnel. In this way, an accurate construction excavation section can be constructed through it, thus ensuring the accuracy of the detection. Therefore, this method is very suitable for large-scale application and promotion. Among them, for example, this method can be run on the overbreak detection side, and optionally, the overbreak detection side can be, but not limited to, a personal computer (PC). It can be understood that the foregoing execution subject does not constitute a limitation on the embodiments of the present application. Correspondingly, the running steps of this method can be, but not limited to, as shown in the following steps S1 to S6.

[0077] S1. Obtain the three-dimensional point cloud dataset of the target construction tunnel and the three-dimensional surface model of the theoretical excavation contour of the target construction tunnel; in this embodiment, for example, the foregoing three-dimensional point cloud dataset is obtained by scanning the target construction tunnel with a laser scanner. Therefore, compared with traditional devices such as total stations and profilers, a large amount of point cloud data of the target construction tunnel can be obtained by one scan in this embodiment without repeated measurement at multiple cross-section positions of the tunnel. Therefore, the detection efficiency can be improved. Further, for example, the three-dimensional surface model of the theoretical excavation contour of the target construction tunnel can be constructed by using the construction design drawings of the target construction tunnel and with the help of software such as 3D Max or Civil 3D. Then, it is stored in the overbreak detection side to facilitate subsequent overbreak detection of the target construction tunnel in combination with the three-dimensional surface model of the actual excavation contour of the target construction tunnel.

[0078] Optionally, in this embodiment, the three-dimensional surface model of the actual excavation section of the tunnel (i.e., the three-dimensional surface model of the actual excavation contour described below) is first constructed according to the collected three-dimensional point cloud dataset. Then, by comparing it with the three-dimensional surface model of the theoretical excavation contour, the overbreak detection result of the tunnel is obtained. The construction process of the actual excavation section model is as shown in the following step S2.

[0079] S2. Based on the three-dimensional point cloud dataset, construct the three-dimensional surface model of the actual excavation contour of the target construction tunnel. Both the three-dimensional surface model of the actual excavation contour and the three-dimensional surface model of the theoretical excavation contour are triangular mesh models. In specific implementation, due to the large number of point cloud datasets collected, to reduce the amount of data to be processed and ensure the construction accuracy of the model at the same time, in this embodiment, the feature point clouds are first extracted, and then, based on the feature point clouds, the model is constructed. At the same time, due to the influence of factors such as equipment, distance, scanning angle, and environment during collection, the point cloud density may be uneven. Therefore, after the model is constructed, in this embodiment, the local density of the feature point clouds is combined to optimize the model, so as to reduce the adverse effects caused by uneven point cloud density (such as reducing holes and gaps, improving the smoothness and accuracy of the surface, etc.).

[0080] Optionally, the foregoing model construction process may but is not limited to the steps shown in S21 to S24 below.

[0081] S21. Extract multiple feature point clouds from the three-dimensional point cloud dataset to form a feature point cloud set. In this embodiment, in order to reduce the number of point clouds while ensuring the construction accuracy of the subsequent excavation section model, in this embodiment, the three-dimensional point cloud dataset is sampled by screening out the feature point clouds. Optionally, this embodiment provides an adaptive voxel sampling method to screen out the feature point clouds from the three-dimensional point cloud set, and the specific implementation process thereof may but is not limited to the steps shown in S21a to S21g below.

[0082] S21a. Initialize the sampling iteration number s, and obtain the side length of the voxel grid at the s-th sampling. When s is 1, the side length of the voxel grid at the s-th time is the initial side length. In specific implementation, the initial side length can be set in advance. After obtaining the side length of the voxel grid at the s-th sampling, each three-dimensional point cloud in the three-dimensional point cloud dataset can be divided into different voxel grids. The division process may but is not limited to the steps shown in S21b below.

[0083] S21b. According to the side length of the voxel grid at the s-th sampling, divide the three-dimensional point cloud dataset into several voxel grids, where each voxel grid contains multiple three-dimensional point clouds. In specific application, for example, but not limited to, first obtain the maximum abscissa, maximum ordinate, maximum z-axis coordinate, minimum abscissa, minimum ordinate, and minimum z-axis coordinate in the three-dimensional point cloud dataset, and then, according to the following formula (3), determine the number of voxel grids divided at the s-th sampling.

[0084] (3)

[0085] In the above formula (3), All represent parameters for calculating the number of voxel grids. Successively represent the maximum abscissa, maximum ordinate, and maximum z - axis coordinate. Successively represent the minimum abscissa, minimum ordinate, and minimum z - axis coordinate. Represents the side length of the voxel grid at the s - th sampling. Represents the ceiling function.

[0086] Therefore, based on the aforementioned formula (3), after calculating the parameters for calculating the number of voxel grids, the product of can be used as the number of voxel grids at the s - th sampling; after calculating the number of voxel grids, each three - dimensional point cloud in the three - dimensional point cloud dataset can be divided into different voxel grids; of course, the process of dividing the three - dimensional point cloud data is a common technique for voxel grid downsampling of point cloud data, and its principle will not be elaborated here.

[0087] After completing the division of the voxel grids at the s - th sampling, sampling processing can be performed on the three - dimensional point clouds in each voxel grid to obtain the feature point clouds in each voxel grid; among them, the selection process of the feature point clouds is as shown in the following step S21c.

[0088] S21c. Extract the feature point clouds at the s - th sampling from multiple three - dimensional point clouds in each voxel grid; in this embodiment, by screening out the neighborhood point clouds of each three - dimensional point cloud in the current voxel grid from the current voxel grid and the voxel grids adjacent to the current voxel grid, and then, by calculating the angles between the normal vectors of each three - dimensional point cloud and the corresponding neighborhood point clouds, to determine whether each three - dimensional point cloud is a feature point cloud, and then extract the feature point clouds.

[0089] Optionally, taking any three - dimensional point cloud in any voxel grid as an example, the process is described as follows, and it can be but is not limited to the following steps S21c1 - S21c6.

[0090] S21c1. For any three - dimensional point cloud in any voxel grid, determine the neighborhood point cloud of the any three - dimensional point cloud from the any voxel grid and the target grid, where the target grid is the voxel grid adjacent to the any voxel grid among the several voxel grids; in specific applications, for example, but not limited to, using k - neighborhood search to search for several neighborhood point clouds corresponding to the any three - dimensional point cloud in the any voxel grid and the target grid.

[0091] After searching for several neighborhood point clouds, the normal vectors of the any three - dimensional point cloud and its corresponding neighborhood point clouds can be calculated, and the process is as shown in the following step S21c2.

[0092] S21c2. Determine the first normal vector of any of the three-dimensional point clouds and the second normal vectors of each neighborhood point cloud; in specific implementation, the covariance matrix of any of the three-dimensional point clouds is constructed by using the neighborhood point clouds corresponding to any of the three-dimensional point clouds, and then the normal vector of any of the three-dimensional point clouds is obtained by solving the eigenvector in the covariance matrix.

[0093] In this embodiment, for example, but not limited to, the following formula (4) can be used to construct the covariance matrix of any of the three-dimensional point clouds.

[0094] (4)

[0095] In the above formula (4), is the covariance matrix of any of the three-dimensional point clouds, represents the centroid coordinates of all neighborhood point clouds corresponding to any of the three-dimensional point clouds, represents the k-th neighborhood point cloud in the neighborhood point clouds of any of the three-dimensional point clouds, represents the total number of neighborhood point clouds of any of the three-dimensional point clouds, represents the transpose operation; and the centroid coordinates are the coordinate average values of all neighborhood point clouds, that is, the average coordinates.

[0096] After constructing the covariance matrix of any of the three-dimensional point clouds based on the foregoing formula (4), the eigenvector in the covariance matrix can be solved; then the smallest eigenvector is used as the normal vector of any of the three-dimensional point clouds; similarly, the process of obtaining the normal vectors of each neighborhood point cloud of any of the three-dimensional point clouds is also the same, which will not be elaborated here.

[0097] After obtaining the normal vectors of any of the three-dimensional point clouds and its corresponding neighborhood point clouds, the angle between the normal vectors can be obtained, and the process is as shown in the following step S21c3.

[0098] S21c3. Calculate the angles between the first normal vector and each second normal vector to obtain multiple normal vector angles; in this embodiment, after calculating the angles between the normal vector of any of the three-dimensional point clouds described above and the normal vectors of its corresponding neighborhood point clouds, the average value of the angles can be calculated, so as to subsequently judge whether any of the three-dimensional point clouds is a feature point cloud based on the average value; among them, the judgment process is as shown in the following steps S21c4 to S21c6.

[0099] S21c4. Obtain the average value of multiple normal vector angles.

[0100] S21c5. Determine whether the average value is greater than or equal to the angle threshold; in this embodiment, the larger the average value of the angles between the normal vectors, the greater the change in the normal vectors within the voxel grid, and the more obvious the feature change of the point cloud; therefore, when the average value of the normal vector angles is greater than or equal to the angle threshold, it can be determined that any three-dimensional point cloud is a feature point cloud in any voxel grid, and the process is as shown in the following step S21c6.

[0101] S21c6. If so, regard any three-dimensional point cloud as a feature point cloud in any voxel grid, and after polling all the three-dimensional point clouds in all voxel grids, obtain the feature point cloud at the s-th sampling.

[0102] In this embodiment, when polling all the three-dimensional point clouds in any voxel grid in the foregoing manner, the feature point cloud in any voxel grid can be extracted. Then, by traversing the remaining voxel grids, the feature point cloud at the s-th sampling can be obtained; finally, it can be determined whether the number of the sampled feature point clouds reaches the preset number, so as to determine whether it is necessary to adjust the side length of the voxel grid to ensure the construction accuracy of the subsequent model while reducing the number of point clouds.

[0103] Among them, the iteration stop judgment condition is as shown in the following step S21d.

[0104] S21d. Determine whether the sampling iteration stop condition is satisfied, where the sampling iteration stop condition is that the number of the feature point clouds at the s-th sampling is equal to the preset number; in this embodiment, the preset number can be specifically set according to actual use and is not specifically limited herein; among them, if the number of the feature point clouds obtained at the s-th sampling is not equal to the preset number, it is necessary to adjust the side length of the voxel grid to adjust the sampling coarseness; the side length adjustment process is as shown in the following step S21e.

[0105] S21e. If not, update the side length of the voxel grid at the s-th sampling according to the number of the feature point clouds at the s-th sampling and the preset number to obtain the updated side length; in specific applications, for example but not limited to, first calculate the voxel grid side length adjustment coefficient at the s-th sampling according to the number of the feature point clouds at the s-th sampling and the preset number; then, calculate the updated side length according to the voxel grid side length adjustment coefficient at the s-th sampling and the initial side length.

[0106] Further, for example but not limited to, use the following formula (1) to calculate the foregoing voxel grid side length adjustment coefficient.

[0107] (1)

[0108] In the above formula (1), represents the voxel grid side length adjustment coefficient at the s-th sampling, Indicates the number of feature point clouds at the sth sampling time, Indicates the preset quantity.

[0109] After calculating the voxel grid side length adjustment coefficient at the s-th sampling, the updated side length can be calculated. In this embodiment, the following formula (2) is used for calculation.

[0110] (2)

[0111] In the above formula (2), represents the updated edge length, represents the initial side length, Represents the control factor; in this embodiment, the control factor is a constant and can be specifically set according to actual use and is not specifically limited here.

[0112] After completing the update of the side length of the voxel grid, the number of voxel grids can be recalculated based on this, and the three-dimensional point cloud can be re-divided, that is, the aforementioned steps S21b to S21e are re-executed until the conditions for stopping the iteration are met, and a feature point cloud set can be obtained; wherein, the iterative sampling process is shown in the following step S21f.

[0113] S21f. Add 1 to s, replace the side length of the voxel grid at the sth sampling with the updated side length, and re-divide the three-dimensional point cloud data set into a number of voxel grids according to the side length of the voxel grid at the sth sampling until the sampling iteration stop condition is met, so as to use the feature point cloud sampled when the sampling iteration stop condition is met to form an initial feature point cloud set.

[0114] In this way, through the aforementioned steps S21b to S21f, this embodiment adjusts the side length of the voxel grid by determining whether the number of feature point clouds obtained by sampling reaches a preset number, and resamples accordingly; and the coarse-grained result of the sampling is affected by the side length of the grid. The larger the side length, the more blurred the geometric information of the grid, and the smaller the side length, the more detailed the local geometric information of the grid. Therefore, this embodiment can balance the retention and simplification of features through adaptive side length adjustment, thereby avoiding insufficient features caused by over-simplification, which causes the problem of low accuracy of subsequent models.

[0115] Therefore, after the initial feature point cloud is extracted, in order to further improve the accuracy of the feature point cloud, this embodiment also performs denoising processing on it, and the process is shown in the following step S21g.

[0116] S21g. Denoise the initial feature point cloud set to obtain the feature point cloud set after denoising. In this embodiment, first, a kd-tree is constructed using the feature point cloud set. Then, the centroid of the feature point cloud set is calculated (the abscissa, ordinate, and z-axis coordinate of the centroid are the mean abscissa, mean ordinate, and mean z-axis coordinate of all initial feature points in the initial feature point cloud set). Next, the average distance from each initial feature point cloud to the centroid is calculated. Then, the standard deviation of the distances from each initial feature point cloud to the centroid is calculated, and based on the standard deviation and the average distance, the search radius is calculated (R = d + α×σ, where α is the radius threshold parameter, σ is the standard deviation, and d is the average distance). Finally, based on the search radius and the distances between each initial feature point cloud and the centroid in the kd-tree, a search is performed in the kd-tree to obtain the feature point cloud set.

[0117] Specifically, the calculated point cloud centroid is used as the reference point, and the search radius R is used as the search range and input into the kd-tree. The kd-tree will retrieve all the initial feature point cloud data located within the search radius range, and thus use them as the feature point cloud to form the feature point cloud set. Of course, using the kd-tree for point cloud search is a common technique for kd-tee filtering, and its specific search process will not be elaborated here.

[0118] Through the foregoing steps S21a to S21g, it is possible to simplify the point cloud data while extracting the feature point cloud that can retain the characteristics of the three-dimensional point cloud data set. Then, based on this, the reconstruction of the excavation section of the target construction tunnel can be carried out, and the process is as shown in the following steps S22 to S24.

[0119] S22. Calculate the local density of each feature point cloud in the feature point cloud set. In specific implementation, as previously described, due to factors such as equipment, distance, scanning angle, and environment during point cloud acquisition, the point cloud density may be uneven. Although point cloud simplification can, to a certain extent, weaken the problem of uneven point cloud density, the feature point cloud set still generally has the problem of uneven density. Therefore, to avoid the adverse effects of uneven point cloud density on subsequent model construction (such as causing holes and gaps, and missing cross-section details), in this embodiment, the local density of the feature point cloud is calculated, and then model optimization is carried out after model construction.

[0120] Optionally, taking any feature point cloud as an example, the calculation process of its local density is described as follows: For any feature point cloud in the feature point cloud set, obtain the k-neighborhood point cloud of the any feature point cloud. First, the distances between the any feature point cloud and each k-neighborhood point cloud can be calculated. Then, the mean of all the distances is used as the local density of the any feature point cloud.

[0121] After calculating the local density of each feature point cloud, the surface reconstruction of the excavation section can be carried out, and the process is as shown in the following step S23.

[0122] S23. Use the set of feature point clouds to construct a three-dimensional surface model of the initial excavation contour of the target construction tunnel; in specific implementation, in this embodiment, an improved Delaunay triangulation algorithm is used to reconstruct the three-dimensional surface of the excavation section of the target construction tunnel, and the process can be but is not limited to the following steps S23a to S23d.

[0123] S23a. Randomly select three feature point clouds from the set of feature point clouds to form an initial triangle; in this embodiment, after randomly selecting three feature point clouds to construct an initial triangle, the longest side can be determined from the initial triangle, and the optimal expansion point cloud corresponding to the longest side can be determined, so that subsequently, based on the optimal expansion point cloud, an expansion triangle can be constructed, and then the expansion triangle can be added to the triangle network corresponding to the initial triangle, thereby completing the expansion of the triangle network; the determination process of the optimal expansion point cloud is as shown in the following step S23b.

[0124] S23b. Take any side of the initial triangle as the growth side, and screen out the optimal expansion point cloud corresponding to the growth side from the set of feature point clouds; in this embodiment, the core of the traditional Delaunay triangulation algorithm is to find the optimal point that meets the conditions. When performing triangulation, it needs to meet the empty circumcircle criterion, that is, for each point found, the circumcircle of this triangle needs to be made, and then it is judged whether there are other points inside the circumcircle. In this way, this irregular selection method will consume a lot of time when the amount of point cloud data is large; therefore, in this embodiment, the search range of the expansion point cloud is limited to reduce the search range, thereby improving the search speed.

[0125] Optionally, for example, but not limited to, the following steps S23b1 to S23b4 can be used to select the optimal expansion point cloud corresponding to the growth side.

[0126] S23b1. Screen out the feature point cloud closest to the growth side from the set of feature point clouds as the initial expansion point cloud; after selecting the feature point cloud closest to the growth side from the set of feature point clouds, an initial expansion triangle can be constructed based on this, as shown in the following step S23b2.

[0127] S23b2. Based on the initial extended point cloud and the growth edge, construct an initial extended triangle; in this embodiment, connect the two ends of the growth edge to the initial extended point cloud respectively, then the initial extended triangle can be obtained. Then, within the range of the circumcircle of the initial extended triangle, search for the optimal extended point cloud, and the process is as shown in the following steps S23b3 to S23b4.

[0128] S23b3. Determine whether there is a specified feature point cloud within the circumcircle of the initial extended triangle, where the specified feature point cloud is all the feature point clouds in the feature point cloud set except the feature point clouds corresponding to the three vertices of the initial extended triangle; in this embodiment, if there is a specified feature point cloud within the circumcircle of the initial extended triangle, randomly select one from the specified feature point clouds as the initial extended point cloud, then reconnect it to the growth edge to re-construct the initial extended triangle; then, re-determine whether there is a specified feature point cloud in the circumcircle of the obtained initial extended triangle. In this way, continuously search according to this principle until there is no specified feature point cloud within the circumcircle. At this time, the initial extended point cloud corresponding to the initial extended triangle is the optimal extended point cloud; the search process is as shown in the following step S23b4.

[0129] S23b4. If so, randomly select one from the specified feature point clouds as the initial extended point cloud, and re-construct an initial extended triangle based on the initial extended point cloud and the growth edge until there is no specified feature point cloud within the circumcircle of the constructed initial extended triangle. Take the initial extended point cloud corresponding to the initial extended triangle without a specified feature point cloud within the circumcircle as the optimal extended point cloud.

[0130] In this embodiment, an example is used to elaborate:

[0131] Suppose the growth edge is AB, and the initial extended point cloud for the first time is the feature point cloud C. Among them, there are five feature point clouds D, E, F, G, and H within the circumcircle of the initial triangle ABC. At this time, randomly select one. Suppose the feature point cloud E is selected. There are three feature point clouds F, G, and H within the circumcircle of the corresponding initial extended triangle ABE. At this time, randomly select one from the three feature point clouds F, G, and H as the initial extended point cloud. Suppose G is selected. There are no other feature point clouds within the circumcircle of the constructed initial extended triangle ABG. Therefore, the feature point cloud G can be used as the optimal extended point cloud of the growth edge AB.

[0132] Thus, through the foregoing description, when triangulating by selecting the optimal expansion point in this embodiment, the selection range is limited to the circumcircle range of the initial expansion triangle constructed by the point cloud closest to the growth edge. In this way, compared with the traditional random selection, the search range is greatly reduced, thereby improving the search efficiency.

[0133] After determining the optimal expansion point cloud corresponding to the growth edge based on the foregoing step S23b and its sub-steps, triangle expansion can be performed based on this, and the steps are as shown in the following step S23c.

[0134] S23c. Use the optimal expansion point cloud and the growth edge to construct an expansion triangle; in this embodiment, connect the two ends of the growth edge with the optimal expansion point cloud respectively, and the expansion triangle can be obtained. For example, based on the above example, the optimal expansion point cloud is the feature point cloud G, then the expansion triangle is triangle ABG. In this way, after one triangulation, the triangular network includes the initial expansion triangle and the expansion triangle ABG; based on this, a new growth edge can be determined on the basis of the expansion triangle, and then repeat the foregoing steps S23b and step S23c to obtain a new expansion triangle. In this way, continuously repeating the foregoing process can complete the three-dimensional triangulation of the feature point cloud set, thereby generating the three-dimensional surface model of the initial excavation contour of the target construction tunnel; among them, the foregoing cyclic triangulation process is as shown in the following step S23d.

[0135] S23d. Update the initial triangle to the expansion triangle, and re-select any side of the initial triangle as the growth edge, and screen out the optimal expansion point cloud corresponding to the growth edge from the feature point cloud set until all the feature point clouds in the feature point cloud set are divided, so as to use the initial triangle and the obtained expansion triangle to construct the three-dimensional surface model of the initial excavation contour.

[0136] Thus, through the foregoing steps S23a to S23d, the surface reconstruction of the excavation section of the target construction tunnel can be completed. Then, the local density of each of the foregoing feature point clouds can be used to optimize the triangular patches, and the process is as shown in the following step S24.

[0137] S24. Optimize the three-dimensional surface model of the initial excavation contour according to the local density of each feature point cloud in the three-dimensional surface model of the initial excavation contour, so as to obtain the three-dimensional surface model of the actual excavation contour after the triangular patch optimization process; in specific applications, an example of its optimization process is as shown in the following steps S24a to S24d.

[0138] S24a. For any initial triangular patch in the three-dimensional surface model of the initial excavation contour, calculate the circumradius of the any initial triangular patch; in this embodiment, for example, but not limited to, sum the lengths of the sides of the any initial triangular patch and divide the sum by 2 to obtain a first intermediate parameter; then, based on the lengths of the three sides of the any initial triangular patch and the first intermediate parameter, calculate a second intermediate parameter; finally, according to the second intermediate parameter and the lengths of the three sides of the any initial triangular patch, calculate the circumradius.

[0139] Optionally, for example, but not limited to, use the following formula (5) to calculate the second intermediate parameter.

[0140] (5)

[0141] In the above formula (5), represents the second intermediate parameter, represents the first intermediate parameter, successively represent the lengths of the three sides of the any initial triangular patch.

[0142] Correspondingly, after calculating the second intermediate parameter, the circumradius can be calculated based on the following formula (6).

[0143] (6)

[0144] In the above formula (6), represents the circumradius.

[0145] After calculating the circumradius of the any initial triangular patch, the any initial triangular patch can be optimized based on the local densities of the feature point clouds corresponding to the three endpoints of the any initial triangular patch, and the process is as shown in the following steps S24b to S24d.

[0146] S24b. According to the local densities of the feature point clouds corresponding to the three endpoints of the any initial triangular patch, calculate the density threshold corresponding to the any initial triangular patch; in this embodiment, for example, but not limited to, first calculate the density mean of the local densities of the feature point clouds corresponding to the three endpoints of the any initial triangular patch; then, multiply the density mean by a density coefficient to obtain the density threshold; optionally, the density coefficient is taken as 0.5.

[0147] After calculating the density threshold corresponding to the any initial triangular patch, it can be combined with its circumradius to determine whether the any initial triangular patch needs to be optimized, and the process is as shown in the following step S24c.

[0148] S24c. Determine whether any of the initial triangular facets is a segmented triangular facet according to the circumradius and the density threshold; in this embodiment, it is determined whether the circumradius is greater than the density threshold to determine whether any of the initial triangular facets needs to be optimized; among them, the larger the circumradius, the larger the spatial range inside the triangle, and the sparser the point cloud distribution (the reason is that a triangle with a large circumradius must also have a large area, which indicates that the distance between the three points of the triangle is far, so it cannot cover the surrounding point cloud well, that is, the point cloud distribution is relatively sparse. At the same time, a triangle with a large circumradius usually has a small interior angle, which means that the triangle is relatively flat and cannot fit the shape of the surrounding surface well, so it is not suitable for surface reconstruction). Therefore, if it is greater than the density threshold, it means that the density of the point cloud distribution in this area is low, which will affect the reconstruction accuracy; therefore, it is necessary to optimize any of the initial triangular facets, that is, insert new points to increase the density to ensure the reconstruction accuracy; otherwise, no optimization is required and it can be retained; optionally, the optimization process is as shown in the following step S24d.

[0149] S24d. If so, insert new feature point clouds into any of the initial triangular facets, so as to perform segmentation processing on any of the initial triangular facets based on the new feature point clouds, and when all the initial triangular facets in the initial excavation contour three-dimensional surface model are polled, obtain the actual excavation contour three-dimensional surface model; in this embodiment, for example, the new feature point clouds are the midpoints of the three sides of any of the initial triangular facets, that is, the midpoints of the three sides are used as the new feature point clouds, and then the original three endpoints are connected, so that any of the initial triangular facets is subdivided into four triangles, so as to achieve the purpose of increasing the density. In this way, according to the foregoing principle, after all the initial triangular facets in the initial excavation contour three-dimensional surface model are optimized, the actual excavation contour three-dimensional surface model of the target construction tunnel can be obtained.

[0150] Thus, through the foregoing steps S21 to S24 and their sub-steps, the actual excavation contour three-dimensional surface model of the target construction tunnel can be constructed by using the collected three-dimensional point cloud data set; then, the under-excavation detection of the tunnel can be carried out in combination with its theoretical excavation contour three-dimensional surface model, and the process is as shown in the following steps S3 to S6.

[0151] S3. For any vertex in the three-dimensional surface model of the actual excavation contour, determine the triangular facet in the three-dimensional surface model of the theoretical excavation contour that is closest to the vertex as the under-excavation detection reference plane for the vertex, and take the distance between the vertex and the under-excavation detection reference plane as the construction error distance. In this embodiment, the distance from any vertex to each triangular facet in the three-dimensional surface model of the theoretical excavation contour is calculated using the distance calculation formula between a point and a plane. Then, the triangular facet with the closest distance is selected as its corresponding under-excavation detection reference plane.

[0152] Based on this, it is possible to determine whether each point on the actual construction excavation section is an under-excavation point according to the positive or negative of the distance from each vertex on the actually constructed three-dimensional surface model of the excavation contour (i.e., the actual construction excavation section) to the corresponding reference plane. The judgment process is as shown in step S4 below.

[0153] S4. Determine whether the construction error distance is 0. In this embodiment, if the distance is 0, it means that the point on the actual excavation section is on the reference plane. At this time, it means that there is no problem of under-excavation or over-excavation. On the contrary, if the distance is not 0, it is necessary to determine whether the vertex is an under-excavation point by judging the positive or negative of the distance. In this embodiment, the sign of the aforementioned distance (i.e., the construction error distance) is determined by the vector of the normal line from the vertex to its corresponding reference plane. The process is as shown in step S5 below.

[0154] S5. If not, determine the sign of the construction error distance according to the vector of the normal line from the vertex to the under-excavation detection reference plane. In this embodiment, assume that the three endpoints of the under-excavation detection reference plane are FHK. Among them, the vector FH can be obtained using the coordinates of points F and H, and the vector FK can be obtained using the coordinates of points F and K. Then, the cross product of the vector FH and the vector FK is calculated to obtain the normal line (i.e., the normal vector) of the under-excavation detection reference plane. Next, the vector from the vertex to the normal line can be calculated. Finally, the dot product of the vector and the normal line can be calculated. If the dot product is greater than 0, it means that the two vectors are in the same direction, and the vertex is actually located inside the reference plane, and the distance is positive. Therefore, it means that the vertex is an over-excavation point. On the contrary, if the dot product result is less than 0, it means that the vertex is actually located outside the reference plane, and the distance is negative, indicating that the vertex is an under-excavation point.

[0155] In this embodiment, the calculation process of the vector from the aforementioned vertex to the normal line is as follows: , where successively represent the vector from the vertex to the normal line, the position vector of the vertex (i.e., the position vector from the vertex to the origin), the unit vector of the normal line of the under-excavation detection reference plane, and Indicates the distance from any one of the vertices to the reference plane.

[0156] Further, the foregoing judgment process is as shown in step S6 below

[0157] S6. If the sign of the construction error distance is negative, it is determined that any one of the vertices is an under-excavated point, and after polling all the vertices in the three-dimensional surface model of the actual excavation contour, the under-excavation detection result of the target construction tunnel is obtained; thus, through the sign of the foregoing construction error distance, the under-excavated points in the three-dimensional surface model of the actual excavation contour can be determined, thereby obtaining the under-excavation detection result of the target construction tunnel; based on this, visualizing it can help construction personnel adjust the construction plan.

[0158] Thus, through the method for detecting under-excavation in high-speed railway tunnel construction detailed in the foregoing steps S1 to S6, the present invention constructs a three-dimensional surface model of the actual excavation contour of the tunnel by collecting the three-dimensional point cloud data set of the target construction tunnel. Then, from the three-dimensional surface model of the theoretical excavation contour of the tunnel, the triangular patch closest to each vertex in the actual model is determined and used as the under-excavation detection reference plane for each vertex; finally, by calculating the positive and negative of the distance from the vertex to the corresponding reference plane, it is determined whether each vertex in the actual model is an under-excavated point, thereby obtaining the final under-excavation detection result; thus, compared with the traditional technology, the three-dimensional point cloud data can be directly scanned by a laser scanner without the need for point-by-point measurement. Therefore, not only the efficiency is improved, but also the measurement workload is reduced; at the same time, a large amount of three-dimensional point cloud data can reflect the actual geometric shape of the tunnel. Thus, an accurate construction excavation section can be constructed through it, thereby ensuring the accuracy of the detection; thus, the present invention is very suitable for large-scale application and promotion.

[0159] As Figure 2 shown, the second aspect of this embodiment provides a hardware device for implementing the method for detecting under-excavation in high-speed railway tunnel construction described in the first aspect of the embodiment, including:

[0160] An acquisition unit for acquiring the three-dimensional point cloud data set of the target construction tunnel and the three-dimensional surface model of the theoretical excavation contour of the target construction tunnel.

[0161] A model construction unit for constructing the three-dimensional surface model of the actual excavation contour of the target construction tunnel based on the three-dimensional point cloud data set, wherein both the three-dimensional surface model of the actual excavation contour and the three-dimensional surface model of the theoretical excavation contour are triangular mesh models.

[0162] An error detection unit, for any vertex in the three-dimensional surface model of the actual excavation contour, is used to determine, from the three-dimensional surface model of the theoretical excavation contour, the triangular patch closest to the any vertex as the under-excavation detection reference plane for the any vertex, and take the distance between the any vertex and the under-excavation detection reference plane as the construction error distance.

[0163] An under-excavation detection unit is used to determine whether the construction error distance is 0.

[0164] An under-excavation detection unit is used to determine the sign of the construction error distance according to the vector of the normal line of the any vertex to the under-excavation detection reference plane when it is determined that the construction error distance is not 0.

[0165] If the sign of the construction error distance is a negative sign, the under-excavation detection unit is further used to determine that the any vertex is an under-excavated point, and after polling all the vertices in the three-dimensional surface model of the actual excavation contour, obtain the under-excavation detection result of the target construction tunnel.

[0166] The working process, working details and technical effects of the device provided in this embodiment can be referred to in the first aspect of the embodiment, and will not be elaborated here.

[0167] As Figure 3 shown, in the third aspect of this embodiment, another under-excavation detection device for high-speed rail tunnel construction is provided. Taking the device as an electronic device as an example, it includes: a memory, a processor and a transceiver that are communicatively connected in sequence, where the memory is used to store a computer program, the transceiver is used to send and receive messages, and the processor is used to read the computer program and execute the under-excavation detection method for high-speed rail tunnel construction as described in the first aspect of the embodiment.

[0168] Specifically, the memory may include, but is not limited to, random access memory (RAM), read only memory (ROM), flash memory, first input first output (FIFO) and / or first in last out (FILO), etc.; specifically, the processor may include one or more processing cores, such as a 4-core processor, an 8-core processor, etc. The processor may be implemented in at least one hardware form of DSP (Digital Signal Processing), FPGA (Field-Programmable Gate Array), PLA (Programmable Logic Array). At the same time, the processor may also include a main processor and a coprocessor. The main processor is a processor used to process data in the wake state, also known as the CPU (Central Processing Unit); the coprocessor is a low-power processor used to process data in the standby state.

[0169] In some embodiments, the processor may be integrated with a GPU (Graphics Processing Unit), and the GPU is responsible for rendering and drawing the content to be displayed on the display screen. For example, the processor may not be limited to using a microprocessor of the STM32F105 series, a reduced instruction set computer (RISC) microprocessor, a processor with an X86 architecture or a processor integrated with an embedded neural-network processing unit (NPU); the transceiver may include, but is not limited to, a Wi-Fi wireless transceiver, a Bluetooth wireless transceiver, a General Packet Radio Service (GPRS) wireless transceiver, a ZigBee (low-power local area network protocol based on the IEEE 802.15.4 standard) wireless transceiver, a 3G transceiver, a 4G transceiver and / or a 5G transceiver, etc. In addition, the device may also include, but is not limited to, a power module, a display screen and other necessary components.

[0170] For the working process, working details and technical effects of the electronic device provided in this embodiment, reference may be made to the first aspect of the embodiment, which will not be elaborated here.

[0171] The fourth aspect of this embodiment provides a storage medium storing instructions for the under-excavation detection method for high-speed railway tunnel construction described in the first aspect of the embodiment, that is, instructions are stored on the storage medium, and when the instructions are run on a computer, the under-excavation detection method for high-speed railway tunnel construction described in the first aspect of the embodiment is executed.

[0172] Among them, the storage medium refers to a carrier for storing data, and may include, but is not limited to, floppy disks, optical discs, hard disks, flash memories, USB flash drives, and / or memory sticks, etc. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices.

[0173] For the working process, working details, and technical effects of the storage medium provided in this embodiment, reference may be made to the first aspect of the embodiment, which will not be elaborated here.

[0174] The fifth aspect of this embodiment provides a computer program product containing instructions, and when the instructions are run on a computer, the computer is made to execute the under-excavation detection method for high-speed railway tunnel construction described in the first aspect of the embodiment. Among them, the computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices.

[0175] Finally, it should be noted that the above are only preferred embodiments of the present invention and are not used to limit the protection scope of the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for detecting under-excavation in high-speed railway tunnel construction, characterized in that, Including: Obtain the three-dimensional point cloud dataset of the target construction tunnel, and the construction design drawings of the target construction tunnel, and construct a three-dimensional surface model of the theoretical excavation contour of the target construction tunnel with the help of 3D Max or Civil 3D software; Based on the three-dimensional point cloud dataset, construct a three-dimensional surface model of the actual excavation contour of the target construction tunnel, where both the actual excavation contour three-dimensional surface model and the theoretical excavation contour three-dimensional surface model are triangular mesh models; For any vertex in the three-dimensional surface model of the actual excavation contour, determine the triangular patch closest to the vertex from the three-dimensional surface model of the theoretical excavation contour as the under-excavation detection reference plane corresponding to the vertex, and use the distance between the vertex and the under-excavation detection reference plane as the construction error distance; Judge whether the construction error distance is 0; If not, determine the sign of the construction error distance according to the vector of the normal line of the vertex to the under-excavation detection reference plane, specifically including: calculate the dot product of the vector and the normal line, if the dot product is greater than 0, determine that the sign of the construction error distance is positive, and if the dot product result is less than 0, determine that the sign of the construction error distance is negative; If the sign of the construction error distance is negative, determine that the vertex is an under-excavated point, and after polling all the vertices in the three-dimensional surface model of the actual excavation contour, obtain the under-excavation detection result of the target construction tunnel.

2. The method according to claim 1, wherein Based on the three-dimensional point cloud dataset, constructing the three-dimensional surface model of the actual excavation contour of the target construction tunnel includes: Extract multiple feature point clouds from the three-dimensional point cloud dataset to form a feature point cloud set, specifically including: initialize the sampling iteration times s, and obtain the side length of the voxel grid at the s-th sampling, where when s is 1, the side length of the voxel grid at the s-th time is the initial side length; divide the three-dimensional point cloud dataset into several voxel grids according to the side length of the voxel grid at the s-th sampling, where each voxel grid contains multiple three-dimensional point clouds; extract the feature point cloud at the s-th sampling from the multiple three-dimensional point clouds in each voxel grid; judge whether the sampling iteration stop condition is satisfied, where the sampling iteration stop condition is that the number of feature point clouds at the s-th sampling is equal to the preset number; if not, update the side length of the voxel grid at the s-th sampling according to the number of feature point clouds at the s-th sampling and the preset number to obtain the updated side length; increment s by 1, and replace the side length of the voxel grid at the s-th sampling with the updated side length, and re-divide the three-dimensional point cloud dataset into several voxel grids according to the side length of the voxel grid at the s-th sampling until the sampling iteration stop condition is satisfied, so as to form an initial feature point cloud set by using the feature point clouds sampled when the sampling iteration stop condition is satisfied; perform denoising processing on the initial feature point cloud set to obtain the feature point cloud set after denoising processing; Calculate the local density of each feature point cloud in the feature point cloud set; Using the set of feature point clouds, construct an initial excavation contour three-dimensional surface model of the target construction tunnel; According to the local density of each feature point cloud in the initial excavation contour three-dimensional surface model, perform triangular facet optimization processing on the initial excavation contour three-dimensional surface model, so as to obtain the actual excavation contour three-dimensional surface model after the triangular facet optimization processing.

3. The method according to claim 2, characterized in that, Extract the feature point cloud at the s-th sampling from the multiple three-dimensional point clouds in each voxel grid, including: For any three-dimensional point cloud in any voxel grid, determine the neighborhood point cloud of the any three-dimensional point cloud from the any voxel grid and the target grid, where the target grid is the voxel grid adjacent to the any voxel grid among the several voxel grids; Determine the first normal vector of the any three-dimensional point cloud and the second normal vectors of each neighborhood point cloud; Calculate the angles between the first normal vector and each second normal vector to obtain a plurality of normal vector angles; Obtain the average value of the plurality of normal vector angles; Judge whether the average value is greater than or equal to the angle threshold; If so, take the any three-dimensional point cloud as a feature point cloud in the any voxel grid, and after polling all the three-dimensional point clouds in all the voxel grids, obtain the feature point cloud at the s-th sampling.

4. The method according to claim 2, characterized in that, Update the side length of the voxel grid at the s-th sampling according to the number of feature point clouds at the s-th sampling and the preset number to obtain the updated side length, including: Calculate the voxel grid side length adjustment coefficient at the s-th sampling according to the number of feature point clouds at the s-th sampling and the preset number; Calculate the updated side length according to the voxel grid side length adjustment coefficient at the s-th sampling and the initial side length.

5. The method according to claim 2, wherein Using the set of feature point clouds, construct an initial excavation contour three-dimensional surface model of the target construction tunnel, including: Randomly select three feature point clouds from the set of feature point clouds to form an initial triangle; Take any side of the initial triangle as the growth side, and screen out the optimal extended point cloud corresponding to the growth side from the set of feature point clouds; Construct an extended triangle using the optimal extended point cloud and the growth side; Update the initial triangle to the extended triangle, and re-take any side of the initial triangle as the growth side, and screen out the optimal extended point cloud corresponding to the growth side from the set of feature point clouds until all the feature point clouds in the set of feature point clouds are divided, so as to construct the initial excavation contour three-dimensional surface model using the initial triangle and the obtained extended triangles.

6. The method according to claim 5, wherein Screen out the optimal extended point cloud corresponding to the growth side from the set of feature point clouds, including: Screen out the feature point cloud closest to the growth side from the set of feature point clouds as the initial extended point cloud; Construct an initial extended triangle based on the initial extended point cloud and the growth side; Judge whether there is a specified feature point cloud inside the circumcircle of the initial extended triangle, where the specified feature point cloud is all the feature point clouds in the set of feature point clouds except the feature point clouds corresponding to the three vertices of the initial extended triangle; If so, randomly select one from the specified feature point clouds as the initial expanded point cloud, and reconstruct the initial expanded triangle based on the initial expanded point cloud and the growth edge until there are no specified feature point clouds within the circumcircle of the constructed initial expanded triangle. Then, use the initial expanded point cloud corresponding to the initial expanded triangle without specified feature point clouds within its circumcircle as the optimal expanded point cloud.

7. The method according to claim 2, wherein Calculate the local density of each feature point cloud in the feature point cloud set, including: For any feature point cloud in the feature point cloud set, obtain the k-neighborhood point cloud of the any feature point cloud; Calculate the distances between the any feature point cloud and each k-neighborhood point cloud, and take the average of all the distances as the local density of the any feature point cloud; Correspondingly, perform triangular patch optimization on the initial excavation contour three-dimensional surface model according to the local densities of the feature point clouds in the initial excavation contour three-dimensional surface model, which includes: For any initial triangular patch in the initial excavation contour three-dimensional surface model, calculate the circumradius of the any initial triangular patch; Calculate the density threshold corresponding to the any initial triangular patch according to the local densities of the feature point clouds corresponding to the three endpoints of the any initial triangular patch; Judge whether the any initial triangular patch is a segmentation triangular patch according to the circumradius and the density threshold; If so, insert a new feature point cloud into the any initial triangular patch, so as to perform segmentation processing on the any initial triangular patch based on the new feature point cloud, and obtain the actual excavation contour three-dimensional surface model when all the initial triangular patches in the initial excavation contour three-dimensional surface model have been polled.

8. A device for detecting under-excavation in high-speed railway tunnel construction, characterized in that, Including: An acquisition unit for acquiring the three-dimensional point cloud dataset of the target construction tunnel, and the construction design drawings of the target construction tunnel, and constructing the theoretical excavation contour three-dimensional surface model of the target construction tunnel with the help of 3D Max or Civil 3D software; A model construction unit for constructing the actual excavation contour three-dimensional surface model of the target construction tunnel based on the three-dimensional point cloud dataset, wherein both the actual excavation contour three-dimensional surface model and the theoretical excavation contour three-dimensional surface model are triangular mesh models; An error detection unit for, for any vertex in the actual excavation contour three-dimensional surface model, determining the triangular patch closest to the any vertex from the theoretical excavation contour three-dimensional surface model as the under-excavation detection reference plane corresponding to the any vertex, and taking the distance between the any vertex and the under-excavation detection reference plane as the construction error distance; An under-excavation detection unit for judging whether the construction error distance is 0; The under-excavation detection unit is used to determine the sign of the construction error distance according to the vector of the normal line from any vertex to the under-excavation detection reference plane when it is determined that the construction error distance is not 0. Specifically, it includes: calculating the dot product of the vector and the normal line. If the dot product is greater than 0, it is determined that the sign of the construction error distance is positive; if the dot product result is less than 0, it is determined that the sign of the construction error distance is negative. If the sign of the construction error distance is negative, the under-excavation detection unit is further used to determine that any vertex is an under-excavation point, and after polling all vertices in the actual excavation contour three-dimensional surface model, the under-excavation detection result of the target construction tunnel is obtained.

9. A computer program product comprising instructions, characterized in that, When the instruction runs on a computer, it causes the computer to perform the high-speed railway tunnel construction under-excavation detection method according to any one of claims 1 to 7.

Images