A method and system for extracting features of tunnel secondary lining steel mesh based on laser point cloud

Abstract

The present invention relates to the field of tunnel secondary lining steel mesh feature recognition, and provides a tunnel secondary lining steel mesh feature extraction method and system based on laser point cloud. The present invention processes the secondary lining steel mesh in steps, firstly obtaining the secondary lining construction data of the target area based on pre-processing such as data correction, segmentation and denoising; secondly separating the front steel mesh based on the data continuity feature; finally separating the rear steel mesh and the waterproof board based on the geometric morphological features of the steel bars, combining the front steel mesh data and traversing the data of each block to obtain complete tunnel secondary lining steel point cloud data, achieving the goal of high-quality extraction of double-layer steel mesh of the tunnel secondary lining. In response to the call for smart transportation construction, the present invention can extract the data of two layers of steel mesh with high precision and automation, improve the reliability of monitoring and detection of the tunnel construction process, and provide technical support for establishing a full life cycle archive of the tunnel structure.

Description

Technical Field

[0001] The present invention relates to the field of tunnel secondary lining steel mesh feature recognition, and in particular to a tunnel secondary lining steel mesh feature extraction method and system based on laser point cloud. Background Art

[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] The secondary lining of a tunnel is the last line of defense for tunnel safety, and the quality of the steel bar construction directly determines the protective performance of the secondary lining. Tunnel engineering has the characteristics of large construction volume, short cycle construction period, and complex and narrow secondary lining construction scene. Therefore, in terms of tunnel engineering quality inspection, there are shortcomings such as low mechanization and automation level, high labor cost, low reliability of sampling inspection, and difficulty in later inspection and calibration.

[0004] In view of the above shortcomings, based on the characteristics of 3D laser point cloud technology, considering the strong adaptability of 3D laser point cloud technology to tunnel scenes, it can achieve large-area, high-precision, full-view and contactless acquisition of coordinate information of tunnel surface points, and then obtain the 3D point cloud data of the second lining construction section structure. However, there are the following shortcomings in accurately extracting the target steel mesh data from the 3D point cloud data: the second lining steel mesh and the tunnel waterproofing board are closely connected, and it is difficult to separate the data; the second layer of steel mesh data is blocked by the front steel bars and the data is incomplete, resulting in the two features are not exactly the same, and it is difficult to extract the two layers of steel mesh uniformly; the amount of point cloud data is huge, the target steel mesh data is secondary data, and the algorithm efficiency is low; there are noise points such as incomplete stirrups and positioning bars between the two layers of steel mesh.

[0005] The existing feature clustering analysis algorithms have various problems in the field of tunnel secondary lining steel mesh feature recognition, and cannot achieve the purpose of complete extraction of the tunnel double-layer steel mesh. Summary of the invention

[0006] In order to solve the technical problems existing in the above-mentioned background technology, the present invention provides a method and system for extracting features of tunnel secondary lining steel mesh based on laser point cloud, which achieves the goal of high-quality extraction of double-layer steel mesh of tunnel secondary lining by processing the secondary lining steel mesh in steps, thereby improving the reliability of monitoring and detection of tunnel construction process, and providing technical support for establishing full life cycle archives of tunnel structure.

[0007] In order to achieve the above object, the present invention adopts the following technical solution:

[0008] A first aspect of the present invention provides a method for extracting features of a tunnel secondary lining steel mesh based on laser point cloud.

[0009] A method for extracting features of tunnel secondary lining steel mesh based on laser point cloud, comprising:

[0010] Obtain tunnel point cloud data and perform preprocessing;

[0011] Traverse the pre-processed tunnel point cloud data, determine the spatial distance between the current point cloud data and the next point cloud data, and if the spatial distance is less than a set threshold, extract the current point cloud data until the traversal is completed to obtain the front side steel mesh feature;

[0012] According to the diameter of the secondary lining steel bar, the neighborhood radius is set, the front steel mesh features are traversed, and the covariance matrix within the neighborhood radius is calculated with the selected current point cloud data as the center of the circle to obtain the eigenvalue of the current point cloud data; the linear length of the eigenvalue of the current point cloud data is calculated and compared with the set linear length threshold, and the point cloud data corresponding to the linear length threshold is extracted until the traversal is completed to obtain the rough extraction features of the rear steel mesh;

[0013] Remove the waterproof board point cloud data from the rough extraction feature of the rear steel mesh to obtain the fine extraction feature of the rear steel mesh;

[0014] The features of the front steel mesh and the rear steel mesh are precisely extracted and arranged according to the set rules to obtain the complete point cloud data of the tunnel secondary lining steel bars.

[0015] Furthermore, the preprocessing includes data correction, data segmentation and encoding, and noise filtering.

[0016] Furthermore, the data correction process includes:

[0017] Through the base points of the tunnel construction site, target points are arranged along the longitudinal direction of the tunnel;

[0018] Use a total station to measure at least two target points and reversely calculate the coordinate information of the installation point;

[0019] Calculate the coordinate transformation matrix according to the coordinate information of the erection point, the coordinate information of the two target points and the three-dimensional coordinates of the tunnel point cloud data;

[0020] According to the coordinate transformation matrix, the three-dimensional coordinates of the tunnel point cloud data are converted into absolute coordinate information.

[0021] Furthermore, the data segmentation and encoding process includes:

[0022] Fit the tunnel centerline, select a fixed length in the tunnel travel direction for slicing, and mark the longitudinal labels;

[0023] Project each segment of the annular second lining along the fitting centerline, perform annular segmentation according to the polar coordinate variable angle, and obtain a two-dimensional data interval with an approximate annular size;

[0024] Each corresponding section of the annular secondary lining is directly divided using this two-dimensional data interval as the boundary to obtain tunnel secondary lining segment data of similar size, and the annular labels are marked in counterclockwise order.

[0025] Furthermore, the noise filtering process includes:

[0026] Traverse the point cloud data, set a search radius r and a minimum number of search points M for each detection point P, and exclude the point if the number of points in the search sphere is less than the minimum number of search points M until the traversal is completed;

[0027] Traverse the point cloud data, set the nearest K points to search for each detection point P, calculate the distance from this point to each point, and calculate the average distance until the traversal is completed;

[0028] Compare K distances. If the distance of a point exceeds twice the average distance, the point is considered to be a discrete point. The points that can be searched directly or indirectly through twice the distance and the point set whose number is less than M' are classified as the same type of discrete data.

[0029] The discrete data labels can only be covered unidirectionally by non-discrete data labels, and finally the discrete data point set is excluded.

[0030] Furthermore, the process of removing the waterproof board point cloud data in the rough extraction feature of the rear steel mesh includes:

[0031] Determine the field radius according to the main rib diameter; traverse the point cloud data and calculate the roughness of each point in the point cloud data until the traversal is completed; exclude the point cloud data corresponding to the roughness greater than the set roughness threshold;

[0032] By setting the minimum number of points within a certain neighborhood radius and excluding isolated noise points, the data on the separation of the steel mesh and the waterproof board can be obtained.

[0033] Furthermore, a plane fitting method is used to fit the noise plane of the waterproof board, excluding points within a certain range from the noise plane, and retaining the loose second layer of steel bar point cloud, that is, the rear side steel bar mesh fine extraction features.

[0034] A second aspect of the present invention provides a tunnel secondary lining steel mesh feature extraction system based on laser point cloud.

[0035] A tunnel secondary lining steel mesh feature extraction system based on laser point cloud, comprising:

[0036] The data acquisition and preprocessing module is configured to: acquire tunnel point cloud data and perform preprocessing;

[0037] The front side steel mesh feature extraction module is configured to: traverse the pre-processed tunnel point cloud data, determine the spatial distance between the current point cloud data and the next point cloud data, and if the spatial distance is less than a set threshold, extract the current point cloud data until the traversal is completed to obtain the front side steel mesh feature;

[0038] The module for rough extraction of the rear steel mesh is configured as follows: according to the diameter of the second lining steel bar, the neighborhood radius is set, the front steel mesh features are traversed, and the covariance matrix within the neighborhood radius is calculated with the selected current point cloud data as the center of the circle to obtain the eigenvalue of the current point cloud data; the linear length of the eigenvalue of the current point cloud data is calculated, and compared with the set linear length threshold, and the point cloud data corresponding to the linear length threshold is extracted until the traversal is completed to obtain the rough extraction features of the rear steel mesh;

[0039] A rear side steel mesh fine extraction module is configured to: remove the waterproof board point cloud data in the rear side steel mesh rough extraction feature to obtain the rear side steel mesh fine extraction feature;

[0040] The output module is configured to: arrange the front side steel mesh features and the rear side steel mesh fine extraction features according to the set rules to obtain complete tunnel secondary lining steel point cloud data.

[0041] A third aspect of the present invention provides a computer-readable storage medium.

[0042] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps in the method for extracting features of a tunnel secondary lining steel mesh based on laser point cloud as described in the first aspect above.

[0043] A fourth aspect of the present invention provides a computer device.

[0044] A computer device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps in the method for extracting features of a tunnel secondary lining steel mesh based on laser point cloud as described in the first aspect above are implemented.

[0045] Compared with the prior art, the present invention has the following beneficial effects:

[0046] The present invention can realize accurate combination of multi-phase secondary lining data through data correction operation, and the extracted steel bar data can be combined into a complete three-dimensional tunnel steel cage point cloud reflecting on-site coordinate information, with good three-dimensional visualization effect.

[0047] The present invention can detect key parts first and then secondary parts according to specifications and requirements through data segmentation and coding operations. First, by selecting key detection parts through segmentation and coding, the detection efficiency can be improved to ensure the normal progress of construction; secondly, most of the existing point cloud algorithms will traverse all points, but the amount of tunnel point cloud data is large, and the total processing time after test data segmentation accounts for about 20% of the complete one-mold two-lining data processing time; finally, after segmentation, the target of feature recognition becomes smaller and the interference factors become less, and some features are more obvious compared to the overall data, thereby improving the extraction accuracy.

[0048] The present invention uses data denoising operation mainly based on the characteristics of tunnel point cloud data to filter out the influence of stirrups, positioning steel bars and isolated noise points to obtain high-precision second lining segment point cloud.

[0049] The present invention uses feature extraction and complementary extraction operations mainly based on the tunnel secondary lining data after the above-mentioned preprocessing operation. It uses the characteristics that the two layers of steel mesh are independent of each other and the distance between the two layers of steel mesh is large after denoising. The two layers of steel mesh are efficiently and accurately separated through distance parameters, and the complete first layer steel mesh data and the second layer steel mesh data of the waterproof board are quickly obtained.

[0050] The present invention extracts the rear steel mesh coarse-fine extraction of features based on the morphological feature that the rear steel mesh is composed of strip steel bars and planar waterproof boards. Therefore, the coarse extraction performs morphological analysis on the point cloud by calculation, excludes most of the planar waterproof board point clouds, reduces the data volume, and greatly improves the computing efficiency; the fine extraction excludes the waterproof board point clouds that occupy the main body of the data by destroying the continuity of the steel bar-waterproof board structure, filtering out detailed noise points, and directional filtering of the waterproof board point clouds, and reversely obtains the looser second layer of steel mesh, combines the complete first layer of steel mesh, and combines the distributed data to finally realize the real-life reproduction of the steel data of the second lining steel construction section, and can check the construction status in real time against the design data.

[0051] The present invention adopts three-dimensional laser scanning technology, which has strong applicability to construction environments, is less affected by external environments such as light, temperature and dust, and is not affected by the presence or absence of signals. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0053] Figure 1 It is the overall flow chart of the feature extraction method of tunnel secondary lining steel mesh based on laser point cloud;

[0054] Figure 2 for Figure 1 Workflow diagram of data segmentation and coding;

[0055] Figure 3 for Figure 1 Noise filtering workflow diagram;

[0056] Figure 4 This is the result diagram of the preprocessing stage in 1;

[0057] Figure 5 for Figure 1 Flow chart of rough extraction of steel bars in the middle and rear sides;

[0058] Figure 6(a) shows Figure 5 Schematic diagram of the theoretical basis of the linear point cloud for rough extraction of the middle and rear side reinforcement;

[0059] Figure 6(b) shows Figure 5 Schematic diagram of the theoretical basis of the surface point cloud for the rough extraction of the steel bars at the middle and rear sides;

[0060] Figure 6(c) shows Figure 5 Schematic diagram of the theoretical basis of the spherical point cloud for the rough extraction of the steel bars at the middle and rear sides;

[0061] Figure 7 for Figure 1 Rough extraction result diagram of the steel bars in the middle and rear sides;

[0062] Figure 8 for Figure 1 Flow chart of fine extraction of steel bars in the middle and rear sides;

[0063] Fig. 9 for Figure 1 Results of the refined extraction of the steel bars in the middle and rear sides;

[0064] Fig.10 This is the overall extraction result diagram of the second lining steel bars. DETAILED DESCRIPTION

[0065] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0066] It should be noted that the following detailed descriptions are all illustrative and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0067] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0068] It should be noted that the flowcharts and block diagrams in the accompanying drawings illustrate the possible implementation architecture, functions and operations of the methods and systems according to various embodiments of the present disclosure. It should be noted that each box in the flowchart or block diagram can represent a module, a program segment, or a part of a code, and the module, program segment, or a part of a code may include one or more executable instructions for implementing the logical functions specified in each embodiment. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in an order different from that marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, or they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the flowchart and / or block diagram, and the combination of boxes in the flowchart and / or block diagram can be implemented using a dedicated hardware-based system that performs a specified function or operation, or can be implemented using a combination of dedicated hardware and computer instructions.

[0069] Embodiment 1

[0070] like Figure 1 As shown, this embodiment provides a method for extracting features of a tunnel secondary lining steel mesh based on laser point cloud. This embodiment uses the method applied to a server as an example. It can be understood that the method can also be applied to a terminal, and can also be applied to a terminal, a server, and a system, and is implemented through the interaction between the terminal and the server. The server can be an independent physical server, or a server cluster or a distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network servers, cloud communications, middleware services, domain name services, security services CDN, and big data and artificial intelligence platforms. The terminal can be a smart phone, a tablet computer, a laptop computer, a desktop computer, a smart speaker, a smart watch, etc., but is not limited to this. The terminal and the server can be directly or indirectly connected via wired or wireless communication, which is not limited in this application. In this embodiment, the method includes the following steps:

[0071] Data correction is used to convert the tunnel point cloud coordinates into geodetic coordinate system coordinates.

[0072] Data segmentation and coding are used to classify the acquired tunnel secondary lining steel bar point cloud data according to spatial location.

[0073] Noise point filtering is used to directionally filter out stirrups, positioning steel bar noise and spatial scattered noise in the secondary lining steel bar point cloud data.

[0074] Front side steel mesh extraction is used to completely extract the data of the first layer of steel mesh in the secondary lining of the tunnel.

[0075] The rough extraction of the rear steel mesh is used to exclude most of the surface waterproofing board point clouds connected to the rear steel mesh.

[0076] The rear steel mesh is precisely extracted to eliminate the connection noise between the waterproof board and the steel bars. The waterproof board noise is eliminated by fitting the plane data of the waterproof board, and the rear steel mesh point cloud with scattered data due to occlusion is retained.

[0077] Specifically, data correction is carried out by arranging target points along the longitudinal direction of the tunnel through the base points of the tunnel construction site, measuring at least two target points using a total station, and reversely calculating the coordinate information of the setting points. It has strong adaptability to complex environments, and inputs the coordinate information of the setting points and the coordinate information of the two target points into the point cloud information, and obtains the three-dimensional coordinates displayed by the point cloud data for calculation, and obtains the coordinate conversion matrix to convert the relative coordinate information into absolute coordinate information.

[0078] like Figure 2 As shown in the figure, data segmentation and encoding are performed by projecting and slicing the secondary lining point cloud data to fit the tunnel centerline, then a fixed length d is selected in the tunnel travel direction for slicing and marking the longitudinal label, and finally each section of the annular secondary lining is projected along the fitting centerline direction, and annular segmentation is performed according to the polar coordinate variable angle to obtain a two-dimensional data interval with similar annular size. Each corresponding section of the annular secondary lining is directly segmented with this two-dimensional data interval as the boundary, thereby obtaining the tunnel secondary lining segment data with similar size and marking the annular labels in counterclockwise order.

[0079] like Figure 3 As shown in the figure, the noise is filtered out. The visual target for the second lining steel bar detection is the steel bar spacing and the steel bar diameter of the steel mesh. The stirrups and positioning steel bars are not within the detection range. Therefore, the steel bar spacing data can be obtained by extracting the two layers of steel mesh and re-measuring the extraction results. The second lining stirrups and limiting steel bars of the tunnel are perpendicular to the two layers of steel mesh and are blocked by the front steel mesh. Therefore, the data of the two obtained by scanning are mostly incomplete, and the continuity between them and the two layers of steel mesh is poor. Based on the above principle, a search radius r and a minimum number of search points M can be set for each detection point P in the process of traversing the point cloud. If the number of points in the search sphere volume is less than the number of points M we set, this point is an isolated noise point and is excluded. Then traverse the point cloud and set the nearest K points to search for each detection point P. Then calculate the distance from this point to each point, find the average distance, and then compare these K distances. If a point is more than twice the average distance away, the point is considered to be a discrete point. The point can be directly or indirectly searched through twice the distance and the number of points is less than M'. The point set is classified as the same type of discrete data. The discrete data label can only be covered by the non-discrete data label in one direction. Finally, the discrete data point set is excluded.

[0080] This embodiment preprocesses the tunnel secondary lining steel construction section data through the above preprocessing process, and can obtain the secondary lining data with independent front and rear side steel meshes and good internal continuity. The results of the preprocessing stage are as follows: Figure 4 Therefore, this embodiment proposes a method for extracting the secondary lining double-layer steel bars based on this.

[0081] The partial extraction of the second lining double-layer steel bars includes the extraction of the front steel mesh, the rough extraction of the rear steel mesh and the fine extraction of the rear steel mesh.

[0082] Among them, the extraction of the front steel mesh refers to the completion of the second lining point cloud data after preprocessing by using the internal continuity distance threshold of the two layers of steel mesh and the difference in the distance between the two layers of steel mesh. The points in the point cloud are traversed. If the distance from the detection point P directly or indirectly to the next point (the next point refers to all the points surrounded by the detection) is less than the set distance threshold M1, these points can be classified as the same type of data until the traversal of the point cloud is completed to achieve different data classifications. According to experimental tests, the value of M1 must be less than the spacing between the two layers of steel bars and greater than the sum of the maximum steel bar diameter Dmax and twice the average point spacing. Among them, the average point spacing can be obtained by extracting a certain number of samples to calculate the distance to the nearest neighbor points around it, and then averaging it.

[0083] Among them, the rough extraction of the rear steel mesh refers to extracting strip steel bars from the remaining data after the above operation, such as Figure 5 As shown in the figure. The point cloud data of the second lining reinforcement construction stage of the tunnel is mainly composed of reinforcement point cloud and waterproof board point cloud. Among them, the waterproof board point cloud is a scattered surface point cloud; the reinforcement segment point cloud is a semi-cylindrical long straight point cloud or a V-surface long straight point cloud, which can be approximated as a linear point cloud. After calculating the covariance matrix of the point cloud data to solve the eigenvalue, the three eigenvalues ​​of λ1, λ2, and λ3 can be obtained from large to small. According to the characteristics of the point cloud neighborhood, the points can be divided into three categories: linear points, surface points, and spherical points. As shown in Figure 6(a), Figure 6(b), and Figure 6(c), the linear point λ1>>λ2≈λ3≈0, the surface point λ1>λ2>>λ3≈0, and the spherical point λ1≈λ2≈λ3.

[0084] Input the target data, set the appropriate neighborhood radius r and linear threshold L' according to the diameter of the secondary lining steel bar, randomly select a point P, calculate the covariance matrix according to the data within the neighborhood radius to solve the three eigenvalues ​​of the point, and calculate the linearity L of the point by the following formula. The larger L is, the stronger the linearity of the point is.

[0085]

[0086] Compare the linear Li of each point with the set linear threshold L'. If Li>L', it is a strip steel bar point cloud, otherwise it is other point clouds. Traverse the point cloud, divide all points into steel bar point clouds and other point clouds, save the steel bar point cloud, end the algorithm, and extract the results as follows Figure 7 shown.

[0087] Among them, the fine extraction of the rear steel mesh refers to selecting the second layer of steel mesh data from the waterproof board data with stronger linearity in the rough extraction results, such as Figure 8 shown.

[0088] Step 1: First, the steel bar point cloud has a certain curvature, but the overall curvature is strong and regular, and the noise data of the waterproof board is relatively discrete; due to the hollowing phenomenon of the waterproof board, the normal distance of the waterproof board data in the middle and at the connection of the steel bar varies greatly from the steel mesh surface; therefore, the curvature of the waterproof board point cloud at the connection varies greatly, and the roughness is stronger. Therefore, according to the diameter of the main bar, a suitable neighborhood radius r is selected, and the point cloud is traversed. The roughness C of each point in the point cloud is calculated, and the points greater than the roughness threshold C' are screened out, so that the connection noise points at the steel mesh and the waterproof board can be excluded. Among them, for each point, its roughness is equal to the distance between the point and the nearest neighbor best fitting plane.

[0089] Step 2: Using density screening, by setting the minimum number of points N within a certain neighborhood radius R, isolated noise points are eliminated to obtain the data of separation of the steel mesh and the waterproof board.

[0090] Step 3: Use the plane fitting method to fit the waterproof board noise plane, exclude points within a certain range from the noise points, and filter out the main noise points of the waterproof board. According to data statistical analysis, the existing mainstream highway tunnel is a three-lane tunnel. When the circumferential longitudinal reinforcement of the steel mesh is about 2m long, the maximum normal bending distance between the middle and the two ends of the steel bar is about 6cm, and the plane curvature is small. In addition, the steel bar data processed above and the waterproof board noise points are manually segmented and their number is counted. It is found that the proportion of steel bars in the total number of points is about 35%. The waterproof board noise points still account for the majority and the continuity between the two has been destroyed through the roughness and density screening process. Therefore, on this basis, the spatial plane of the waterproof board noise points can be fitted and the points whose point-to-plane distance is less than the distance threshold D' can be screened out, thereby retaining the loose second layer of steel bar point cloud. The results are as follows: Fig. 9 shown.

[0091] Finally, the separated first layer steel bar point cloud and the extracted second layer steel bar point cloud are arranged according to the rules in the data segmentation and encoding process, and finally the complete tunnel secondary lining steel bar point cloud data is obtained, such as Fig.10 By comparing the recall rate and precision of the change in the number of steel bar point clouds before and after separation and performing secondary calculations, the extraction rate of the first layer of steel mesh by this method is 100%, the extraction rate of the second layer of steel bars is 60%, and the comprehensive extraction rate is 80%.

[0092] Embodiment 2

[0093] This embodiment provides a tunnel secondary lining steel mesh feature extraction system based on laser point cloud.

[0094] A tunnel secondary lining steel mesh feature extraction system based on laser point cloud, comprising:

[0095] The data acquisition and preprocessing module is configured to: acquire tunnel point cloud data and perform preprocessing;

[0096] The front side steel mesh feature extraction module is configured to: traverse the pre-processed tunnel point cloud data, determine the spatial distance between the current point cloud data and the next point cloud data, and if the spatial distance is less than a set threshold, extract the current point cloud data until the traversal is completed to obtain the front side steel mesh feature;

[0097] The module for rough extraction of the rear steel mesh is configured as follows: according to the diameter of the second lining steel bar, the neighborhood radius is set, the front steel mesh features are traversed, and the covariance matrix within the neighborhood radius is calculated with the selected current point cloud data as the center of the circle to obtain the eigenvalue of the current point cloud data; the linear length of the eigenvalue of the current point cloud data is calculated, and compared with the set linear length threshold, and the point cloud data corresponding to the linear length threshold is extracted until the traversal is completed to obtain the rough extraction features of the rear steel mesh;

[0098] A rear side steel mesh fine extraction module is configured to: remove the waterproof board point cloud data in the rear side steel mesh rough extraction feature to obtain the rear side steel mesh fine extraction feature;

[0099] The output module is configured to: arrange the front side steel mesh features and the rear side steel mesh fine extraction features according to the set rules to obtain complete tunnel secondary lining steel point cloud data.

[0100] It should be noted that the above data acquisition and preprocessing module, front steel mesh feature extraction module, rear steel mesh rough extraction module, rear steel mesh fine extraction module and output module are the same as the examples and application scenarios implemented by the steps in Embodiment 1, but are not limited to the contents disclosed in Embodiment 1. It should be noted that the above modules as part of the system can be executed in a computer system such as a set of computer executable instructions.

[0101] Embodiment 3

[0102] This embodiment provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the steps in the method for extracting features of a tunnel secondary lining steel mesh based on laser point cloud as described in the first embodiment above are implemented.

[0103] Embodiment 4

[0104] This embodiment provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the steps in the method for extracting features of a tunnel secondary lining steel mesh based on laser point cloud as described in the first embodiment are implemented.

[0105] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of hardware embodiments, software embodiments, or embodiments combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage and optical storage, etc.) containing computer-usable program code.

[0106] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0107] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0108] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable device. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0109] A person skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium, and when the program is executed, it can include the processes of the embodiments of the above-mentioned methods. The storage medium can be a disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM), etc.

[0110] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for extracting features of tunnel secondary lining steel mesh based on laser point cloud, characterized in that: include: Obtain tunnel point cloud data and perform preprocessing; Traverse the pre-processed tunnel point cloud data, determine the spatial distance between the current point cloud data and the next point cloud data, and if the spatial distance is less than a set threshold, extract the current point cloud data until the traversal is completed to obtain the front side steel mesh feature; According to the diameter of the secondary lining steel bar, the neighborhood radius is set, the front steel mesh features are traversed, and the covariance matrix within the neighborhood radius is calculated with the selected current point cloud data as the center of the circle to obtain the eigenvalue of the current point cloud data; Calculate the linear length of the characteristic value of the current point cloud data, and compare it with the set linear length threshold, extract the point cloud data corresponding to the linear length threshold, until the traversal is completed, and obtain the rough extraction features of the rear steel mesh; Remove the waterproof board point cloud data from the rough extraction feature of the rear steel mesh to obtain the fine extraction feature of the rear steel mesh; The front steel mesh features and the rear steel mesh fine extraction features are arranged according to the set rules to obtain the complete tunnel secondary lining steel point cloud data; The preprocessing includes data correction, data segmentation and encoding, and noise filtering; The data correction process includes: Through the base points of the tunnel construction site, target points are arranged along the longitudinal direction of the tunnel; Use a total station to measure at least two target points and reversely calculate the coordinate information of the installation point; Calculate the coordinate transformation matrix according to the coordinate information of the erection point, the coordinate information of the two target points and the three-dimensional coordinates of the tunnel point cloud data; According to the coordinate conversion matrix, the three-dimensional coordinates of the tunnel point cloud data are converted into absolute coordinate information; The data segmentation and encoding process includes: Fit the tunnel centerline, select a fixed length in the tunnel travel direction for slicing, and mark the longitudinal labels; Project each segment of the annular second lining along the fitting centerline, perform annular segmentation according to the polar coordinate variable angle, and obtain a two-dimensional data interval with an approximate annular size; Use this two-dimensional data interval as the boundary to directly segment each corresponding circumferential secondary lining, obtain tunnel secondary lining segment data of similar size, and mark circumferential labels in counterclockwise order; The noise filtering process includes: Traverse the point cloud data, set a search radius r and a minimum number of search points M for each detection point P, and exclude the point if the number of points in the search sphere is less than the minimum number of search points M until the traversal is completed; Traverse the point cloud data, set the nearest K points to search for each detection point P, calculate the distance from this point to each point, and calculate the average distance until the traversal is completed; Compare K distances. If the distance of a point exceeds twice the average distance, the point is considered to be a discrete point. The points that can be searched directly or indirectly through twice the distance and the point set whose number is less than M' are classified as the same type of discrete data. The discrete data labels can only be covered unidirectionally by non-discrete data labels, and finally the discrete data point set is excluded.

2. The method for extracting features of tunnel secondary lining steel mesh based on laser point cloud according to claim 1 is characterized in that: The process of removing the waterproof board point cloud data in the rough extraction feature of the rear steel mesh includes: Determine the field radius according to the main rib diameter; traverse the point cloud data and calculate the roughness of each point in the point cloud data until the traversal is completed; exclude the point cloud data corresponding to the roughness greater than the set roughness threshold; By setting the minimum number of points within a certain neighborhood radius and excluding isolated noise points, the data on the separation of the steel mesh and the waterproof board can be obtained.

3. The method for extracting features of tunnel secondary lining steel mesh based on laser point cloud according to claim 2 is characterized in that: The plane fitting method is used to fit the noise plane of the waterproof board, and the points within a certain range from the noise plane are excluded, and the loose second layer of steel bar point cloud, that is, the rear side steel bar mesh is precisely extracted.

4. A tunnel secondary lining steel mesh feature extraction system based on laser point cloud, characterized in that: include: The data acquisition and preprocessing module is configured to: acquire tunnel point cloud data and perform preprocessing; The preprocessing includes data correction, data segmentation and encoding, and noise filtering; The data correction process includes: Through the base points of the tunnel construction site, target points are arranged along the longitudinal direction of the tunnel; Use a total station to measure at least two target points and reversely calculate the coordinate information of the installation point; Calculate the coordinate transformation matrix according to the coordinate information of the erection point, the coordinate information of the two target points and the three-dimensional coordinates of the tunnel point cloud data; According to the coordinate conversion matrix, the three-dimensional coordinates of the tunnel point cloud data are converted into absolute coordinate information; The data segmentation and encoding process includes: Fit the tunnel centerline, select a fixed length in the tunnel travel direction for slicing, and mark the longitudinal labels; Project each segment of the annular second lining along the fitting centerline, perform annular segmentation according to the polar coordinate variable angle, and obtain a two-dimensional data interval with an approximate annular size; Use this two-dimensional data interval as the boundary to directly segment each corresponding circumferential secondary lining, obtain tunnel secondary lining segment data of similar size, and mark circumferential labels in counterclockwise order; The noise filtering process includes: Traverse the point cloud data, set a search radius r and a minimum number of search points M for each detection point P, and exclude the point if the number of points in the search sphere is less than the minimum number of search points M until the traversal is completed; Traverse the point cloud data, set the nearest K points to search for each detection point P, calculate the distance from this point to each point, and calculate the average distance until the traversal is completed; Compare K distances. If the distance of a point exceeds twice the average distance, the point is considered to be a discrete point. The points that can be searched directly or indirectly through twice the distance and the point set whose number is less than M' are classified as the same type of discrete data. The discrete data label can only be covered by the non-discrete data label in one direction, and finally the discrete data point set is excluded; The front side steel mesh feature extraction module is configured to: traverse the pre-processed tunnel point cloud data, determine the spatial distance between the current point cloud data and the next point cloud data, and if the spatial distance is less than a set threshold, extract the current point cloud data until the traversal is completed to obtain the front side steel mesh feature; The module for rough extraction of the rear steel mesh is configured as follows: according to the diameter of the second lining steel bar, the neighborhood radius is set, the front steel mesh features are traversed, and the covariance matrix within the neighborhood radius is calculated with the selected current point cloud data as the center of the circle to obtain the eigenvalue of the current point cloud data; the linear length of the eigenvalue of the current point cloud data is calculated, and compared with the set linear length threshold, and the point cloud data corresponding to the linear length threshold is extracted until the traversal is completed to obtain the rough extraction features of the rear steel mesh; A rear side steel mesh fine extraction module is configured to: remove the waterproof board point cloud data in the rear side steel mesh rough extraction feature to obtain the rear side steel mesh fine extraction feature; The output module is configured to: arrange the front side steel mesh features and the rear side steel mesh fine extraction features according to the set rules to obtain complete tunnel secondary lining steel point cloud data.

5. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps in the method for extracting features of a tunnel secondary lining steel mesh based on laser point cloud as described in any one of claims 1 to 3 are implemented.

6. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the steps in the method for extracting features of tunnel secondary lining steel mesh based on laser point cloud as described in any one of claims 1 to 3 are implemented.

Images