A tunnel LiDAR point cloud processing and deformation detection method and device

Abstract

The present application relates to the technical field of tunnel structure health detection, and particularly relates to a tunnel LiDAR point cloud processing and deformation detection method and device. The method comprises point cloud data preprocessing: loading two-period LiDAR point cloud data of the tunnel, automatically denoising the point cloud data by using a DBSCAN algorithm, and manually reserving target area point cloud data by frame selection denoising; reference correction: extracting a tunnel construction joint as a control point, and importing control point data; constructing a rough path line, the rough path line comprising a first rough path line and a second rough path line, calculating a first-period along-line distance based on the first rough path line, calculating a second-period along-line distance based on the second rough path line, calculating a corrected along-line distance of the second-period control point based on the first-period along-line distance and the second-period along-line distance, and completing two-period point cloud rough reference alignment based on the corrected along-line distance. In this way, the point cloud offset problem in the tunnel signal-free scene is solved, the central axis fitting accuracy is improved, and multi-dimensional accurate deformation detection is achieved.

Description

Technical Field

[0001] This invention relates to the field of tunnel structure health detection technology, specifically to a method and apparatus for tunnel LiDAR point cloud processing and deformation detection. Background Technology

[0002] Tunnel engineering is a core infrastructure of the transportation industry, widely used in highways, railways, and other fields. Its structural health directly affects the safety and stability of transportation. With the expansion of tunnel construction scale and the increase in operating years in my country, structural defects such as lining deformation, road settlement, and maintenance lane misalignment are becoming increasingly prominent, placing higher demands on tunnel inspection technology.

[0003] Currently, tunnel inspection mainly relies on traditional methods such as manual measurement and optical instrument inspection, which suffers from pain points such as low inspection efficiency, high labor intensity, limited inspection accuracy, and cumbersome data processing. With the development of LiDAR technology, LiDAR point cloud detection has been gradually applied to the field of tunnel inspection due to its high efficiency and comprehensiveness. However, existing LiDAR point cloud processing technology still has many shortcomings: in tunnel no-signal scenarios, the two phases of point cloud data are prone to offset errors, resulting in low point cloud alignment accuracy; the centerline fitting uses traditional B-spline curves, which are difficult to adapt to the curved contour of the tunnel, resulting in insufficient fitting accuracy; deformation detection is mostly single-dimensional, which cannot fully cover the deformation of the lining, pavement, and maintenance road, and the anomaly judgment is not accurate enough; at the same time, the operation of existing systems is complex, requiring high professional skills from operators, and it is difficult to meet the high-efficiency inspection needs of engineering sites.

[0004] Therefore, developing a tunnel LiDAR point cloud processing and deformation detection method and device that can solve the above-mentioned technical pain points, achieve accurate point cloud alignment, high-precision centerline fitting, multi-dimensional deformation detection, and is easy to operate has become an urgent need in the current tunnel inspection field. Summary of the Invention

[0005] The purpose of this invention is to provide a method and apparatus for tunnel LiDAR point cloud processing and deformation detection. Through core algorithm optimization and module design, it improves the accuracy of point cloud alignment and deformation detection, simplifies the operation process, adapts to tunnel scenarios without signal, and meets the needs of efficient detection in engineering sites.

[0006] In a first aspect of the present invention, a method for tunnel LiDAR point cloud processing and deformation detection is provided, comprising:

[0007] S1. Point cloud data preprocessing: Load the LiDAR point cloud data of the two phases of the tunnel, use the DBSCAN algorithm to automatically denoise the point cloud data, and manually retain the point cloud data of the target area by selecting the area for denoising.

[0008] S2. Benchmark Correction: Extract tunnel construction joints as control points and import control point data; construct coarse path lines, including a first coarse path line and a second coarse path line; calculate the first phase distance along the line based on the first coarse path line; calculate the second phase distance along the line based on the second coarse path line; calculate the corrected distance along the line for the second phase control points based on the first phase distance and the second phase distance; and complete the coarse benchmark alignment of the two phase point clouds based on the corrected distance.

[0009] S3. Centerline Fitting: Generate tunnel cross-section, obtain centroid coordinates based on tunnel cross-section, and generate cross-section contour line based on centroid coordinates; extract side vertical point clouds of maintenance lanes on both sides, obtain boundary lines of maintenance lanes on both sides based on side vertical point clouds of maintenance lanes on both sides, obtain local center Xf and spacing XL of boundary lines on both sides, filter out qualified tunnel cross-sections that do not exceed the first preset threshold based on spacing XL, and generate tunnel centerline based on centroid coordinates and local center Xf coordinates of qualified tunnel cross-sections.

[0010] S4. Deformation Detection: Deformation detection of the tunnel lining, the tunnel pavement, and the tunnel maintenance road.

[0011] S5. Data Export: Export tunnel section information, deformation detection results, and centerline information.

[0012] As a preferred embodiment of the present invention, in S1, the point cloud data is in LAS 1.2 / 1.3 format; the automatic denoising of the point cloud data specifically involves setting the neighborhood radius and the minimum number of neighborhood points to remove outliers and invalid points in the point cloud data; the manual retention of the target area point cloud data through selection denoising specifically involves selecting the target area with the mouse to remove the point cloud data outside the selection.

[0013] In a preferred embodiment of the present invention, in S2, importing control point data specifically includes importing control point data containing control point numbers, three-dimensional coordinates, and the phase to which they belong. The control point data includes first-phase control points and second-phase control points. Constructing a coarse path line specifically includes constructing a first coarse path line and a second coarse path line based on the path line of the first lane of the tunnel, either by manually drawing or by importing an SHP vector file. Calculating the first-phase distance along the path based on the first coarse path line specifically includes projecting the first-phase control points onto the first-phase coarse path line and calculating the length of the projection point from the path start point to obtain the first-phase distance along the path for the first-phase control points. Calculating the second-phase distance along the path based on the second coarse path line specifically includes projecting the second-phase control points onto the second-phase coarse path line and calculating the length of the projection point from the path start point to obtain the second-phase distance along the path for the second-phase control points.

[0014] As a preferred embodiment of the present invention, the corrected distance along the line for the second-phase control point is calculated using the following formula:

[0015]

[0016] Where is the corrected distance along the line for the second phase control point, is the second phase distance along the line for the second phase control point, and is the first phase distance along the line for the first phase control point.

[0017] In a preferred embodiment of the present invention, in S3, generating the tunnel cross section specifically includes generating the tunnel cross section based on the coarse path line and according to the preset cross section spacing and cross section width; obtaining the centroid coordinates based on the tunnel cross section specifically includes projecting the point cloud within the tunnel cross section onto a projection plane perpendicular to the coarse path line, calling the DBSCAN algorithm to remove outliers in the projection plane, and calculating the centroid coordinates of the effective point cloud in the projection plane; generating the cross section contour line based on the centroid coordinates specifically includes using an improved B-spline curve fitting algorithm and generating the cross section contour line based on the centroid coordinates; extracting the side vertical point cloud of the maintenance lanes on both sides specifically includes extracting the side vertical point cloud of the maintenance lanes using a histogram algorithm; obtaining the boundary lines of the maintenance lanes on both sides based on the side vertical point cloud of the maintenance lanes on both sides specifically includes removing extreme points and then using a least squares algorithm to fit the boundary lines of the left and right maintenance lanes.

[0018] As a preferred embodiment of the present invention, in S4, the lining deformation detection of the tunnel section specifically includes: dividing the projection plane of the first-phase tunnel section into several regions according to an angle, with the central axis of the first-phase tunnel as the center, and obtaining the distance from the first-phase outline of each region to the center; dividing the projection plane of the second-phase tunnel section into several regions according to an angle, with the central axis of the second-phase tunnel as the center, and obtaining the distance from the second-phase outline of each region to the center; calculating the difference between the distances of the corresponding regions as the lining deformation amount, determining whether the lining deformation amount exceeds a second preset threshold, and marking the region as lining abnormal when the lining deformation amount of a certain region exceeds the second preset threshold;

[0019] The specific steps for detecting tunnel pavement deformation include: dividing the first-phase tunnel into segments and calculating the first-phase pavement elevation for each segment; dividing the second-phase tunnel into segments and calculating the second-phase pavement elevation for each segment; calculating the difference between the first-phase and second-phase pavement elevations for the corresponding segment as the pavement deformation amount; determining whether the pavement deformation amount exceeds a third preset threshold; and marking the segment as pavement abnormal when the pavement deformation amount of a certain segment exceeds the third preset threshold.

[0020] The specific steps for detecting the deformation of the tunnel section include: obtaining the local center Xf and the highest elevation point of the first phase for each first-phase tunnel section; obtaining the local center Xf and the highest elevation point of the second phase for each second-phase tunnel section; calculating the difference between the local center Xf of the first phase and the local center Xf of the second phase for the corresponding tunnel section as the deformation in the X direction of the inspection road; calculating the difference between the highest elevation point of the first phase and the highest elevation point of the second phase for the corresponding tunnel section as the deformation in the Y direction of the inspection road; determining whether the deformation in the X direction of the inspection road exceeds the fourth preset threshold and whether the deformation in the Y direction of the inspection road exceeds the fifth preset threshold; and marking the tunnel section as having an abnormal inspection road when the deformation in the X direction of the inspection road of a certain tunnel section exceeds the fourth preset threshold or the deformation in the Y direction of the inspection road exceeds the fifth preset threshold.

[0021] As a preferred embodiment of the present invention, in S5, exporting the tunnel cross-section information, deformation detection results, and centerline information specifically includes exporting the tunnel cross-section information, deformation detection results, and centerline information in CAD, SHP vector, Excel report, and TXT analysis report formats, respectively.

[0022] In a second aspect of the present invention, a tunnel LiDAR point cloud processing and deformation detection device is provided, comprising:

[0023] The point cloud data preprocessing module is configured to load LiDAR point cloud data from the two phases of the tunnel, use the DBSCAN algorithm to automatically denoise the point cloud data, and manually retain the point cloud data of the target area through box selection denoising.

[0024] The benchmark correction module is configured to extract tunnel construction joints as control points, import control point data, construct coarse path lines, including a first coarse path line and a second coarse path line, calculate the first phase distance along the line based on the first coarse path line, calculate the second phase distance along the line based on the second coarse path line, and calculate the corrected distance along the line of the second phase control points based on the first phase distance and the second phase distance.

[0025] The central axis fitting module is configured to generate a tunnel cross section, obtain the centroid coordinates based on the tunnel cross section, generate the cross section outline based on the centroid coordinates, extract the side vertical point cloud of the maintenance lanes on both sides, obtain the boundary lines of the maintenance lanes on both sides based on the side vertical point cloud of the maintenance lanes on both sides, obtain the local center Xf and spacing XL of the boundary lines on both sides, filter out qualified tunnel cross sections that do not exceed the first preset threshold based on the spacing XL, and fit the centroid coordinates and local center Xf coordinates of the qualified tunnel cross sections to generate the tunnel central axis.

[0026] The deformation detection module is configured to detect the deformation of the tunnel lining, the tunnel pavement, and the tunnel maintenance road.

[0027] The data export module is configured to export tunnel cross-section information, deformation detection results, and centerline information.

[0028] In a third aspect of the present invention, a computer program product is provided, including a computer program that, when executed by a processor, implements the method provided according to the first aspect.

[0029] In a fourth aspect of the present invention, an electronic device is provided, including one or more processors and a memory associated with the one or more processors, the memory being used to store program instructions that, when read and executed by the one or more processors, perform the method provided in the first aspect.

[0030] In summary, the present invention has the following beneficial effects:

[0031] The problem of point cloud offset in tunnel no-signal scenarios has been solved: by using a dynamic distance correction algorithm, based on the distance difference compensation between the two control points along the line, the two phases of point clouds are accurately aligned, which greatly improves the point cloud processing accuracy and avoids the impact of offset error on the detection results;

[0032] Improved accuracy of centerline fitting: An improved B-spline curve fitting algorithm was adopted, combined with centroid positioning and inspection road parameter screening, abnormal sections were eliminated, and the generated centerline more closely matches the actual direction and contour center of the tunnel, providing a reliable benchmark for deformation detection;

[0033] Achieve multi-dimensional and accurate deformation detection: covering three dimensions: lining, pavement, and maintenance road. Lining deformation is detected by angle zone, pavement deformation is based on elevation difference analysis, and maintenance road offset is calculated by two-way difference. Anomaly judgment is accurate and comprehensively reflects the health status of the tunnel structure.

[0034] Easy to operate and highly efficient: The operation process is optimized, a visual interactive interface is provided, core functions can be triggered with one click, batch processing and multi-format data export are supported, reducing the professional requirements of operators and adapting to the high-efficiency testing needs of engineering sites.

[0035] Highly practical and widely adaptable: It supports importing and exporting data in multiple formats, is compatible with CAD, GIS, Excel and other software, and is suitable for structural health monitoring of various tunnels such as highways and railways, making it highly valuable for widespread application.

[0036] Further or more detailed beneficial effects will be described in conjunction with specific embodiments in the detailed implementation. Attached Figure Description

[0037] Figure 1 A flowchart of the tunnel LiDAR point cloud processing and deformation detection method according to an embodiment of the present invention is shown;

[0038] Figure 2A block diagram of a tunnel LiDAR point cloud processing and deformation detection device according to an embodiment of the present invention is shown;

[0039] Figure 3 A block diagram of an electronic device according to an embodiment of the present invention is shown;

[0040] Figure 4 A schematic diagram of the cross-sectional profile of an embodiment of the present invention is shown. Detailed Implementation

[0041] Embodiments of the present invention will now be described in more detail with reference to the accompanying drawings. While some embodiments of the invention are shown in the drawings, it should be understood that the invention can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of the invention. It should be understood that the accompanying drawings and embodiments are for illustrative purposes only and are not intended to limit the scope of protection of the invention.

[0042] In the description of embodiments of the present invention, the term "comprising" and similar terms should be understood as open-ended inclusion, i.e., "including but not limited to". The term "based on" should be understood as "at least partially based on". The term "one embodiment" or "the embodiment" should be understood as "at least one embodiment". The terms "first", "second", etc., may refer to different or the same objects. Other explicit and implicit definitions may also be included below.

[0043] Figure 1 A flowchart of a tunnel LiDAR point cloud processing and deformation detection method according to an embodiment of the present invention is shown. The method includes:

[0044] S1. Point cloud data preprocessing: Load the LiDAR point cloud data from the two phases of the tunnel, use the DBSCAN algorithm to automatically denoise the point cloud data, and manually retain the point cloud data of the target area by selecting the area for denoising.

[0045] In this embodiment, the point cloud data is in LAS 1.2 / 1.3 format. The staff first imports the original LAS 1.2 / 1.3 format point cloud files obtained from the two phases of scanning into the point cloud processing software. At this point, hundreds of millions of densely packed laser points will be displayed on the computer screen, forming the three-dimensional shape of the tunnel.

[0046] In this embodiment, automatic denoising of point cloud data specifically involves setting a neighborhood radius and a minimum number of neighboring points to remove outliers and invalid points from the point cloud data. In this embodiment, the default neighborhood radius of the DBSCAN algorithm is 0.5m, and the default minimum number of neighboring points is 10, which can be adjusted according to the point cloud density. The DBSCAN algorithm traverses each point like "checking a household registration." If a point is within a 0.5-meter radius and cannot find at least 10 "neighbors" adjacent to it, the algorithm will determine that it is an isolated noise point and automatically remove it. After this step, all the scattered invalid noise points floating outside the tunnel outline are removed, leaving only the tightly clustered, valid point cloud that constitutes the tunnel entity.

[0047] In this embodiment, manually retaining the point cloud data of the target area through bounding selection denoising specifically involves selecting the target area with a mouse to remove point cloud data outside the selected area, thus avoiding accidental deletion of valid point cloud data. To ensure the integrity of tunnel scanning, field data acquisition often involves "scanning more," for example, the scanner might scan tens of meters of mountain rocks behind the tunnel lining, or even the road outside the tunnel entrance. In the point cloud processing software, staff can switch to a top-down or side-view perspective and use the "polygonal bounding tool" to draw a closed frame along the actual contour of the tunnel, only framing the tunnel lining, road surface, and maintenance access area, deleting all useless mountain rocks and street point cloud data outside the frame. This ultimately yields a clean, pure, high-quality point cloud data containing only the tunnel itself, preparing for subsequent centerline fitting and deformation detection.

[0048] S2. Benchmark Correction: Extract tunnel construction joints as control points and import control point data; construct coarse path lines, including a first coarse path line and a second coarse path line; calculate the first phase distance along the line based on the first coarse path line; calculate the second phase distance along the line based on the second coarse path line; calculate the corrected distance along the line for the second phase control points based on the first and second phase distances along the line; and complete the coarse benchmark alignment of the two phase point clouds based on the corrected distances along the line.

[0049] In this embodiment, importing control point data specifically includes importing control point data containing control point numbers, three-dimensional coordinates, and the period to which they belong. The control point data includes control points from the first period and control points from the second period.

[0050] Specifically, for the first phase of control points, staff accessed the tunnel point cloud for the first phase. First, they located the 10th ring construction joint, clicked the center of the joint with the mouse, and the point cloud processing software recorded the 3D coordinates of this point. A pop-up window then prompted the staff to enter the number "10". At this point, a control point belonging to the first phase was successfully "extracted" and "imported" into the database. Next, they located the 20th ring construction joint, clicked the center of the joint with the mouse, recorded the 3D coordinates of this point, and a pop-up window prompted the staff to enter the number "20". Again, a control point belonging to the first phase was successfully "extracted" and "imported" into the database. By repeating this process, the required number of first-phase control points could be extracted and imported.

[0051] The same procedure applies to the second phase of control points. The staff opens the tunnel point cloud for the second phase. First, they locate the 10th ring construction joint. The staff clicks the center of the joint with the mouse, and the point cloud processing software records the 3D coordinates of this point. A pop-up window prompts the staff to enter the number "10". At this point, a control point belonging to the second phase is successfully "extracted" and "imported" into the database. Next, they locate the 20th ring construction joint. The staff clicks the center of the joint with the mouse, and the point cloud processing software records the 3D coordinates of this point. A pop-up window prompts the staff to enter the number "20". Again, a control point belonging to the second phase is successfully "extracted" and "imported" into the database. By repeating this process, the required number of second-phase control points can be extracted and imported.

[0052] Import at least three control points to ensure the reliability of the benchmark calibration.

[0053] In this embodiment, constructing the coarse path line specifically includes using the path line of the first lane of the tunnel as a reference, and constructing the first coarse path line and the second coarse path line by manually drawing or importing an SHP vector file.

[0054] Specifically, for the first coarse path line, the staff opened the first phase of the tunnel point cloud, switched the view to a top view, and used the "polyline drawing tool" in the point cloud processing software to click the mouse at intervals (e.g., 10 or 20 meters) along the center of the first lane of the tunnel. Finally, the point cloud processing software automatically connected these points with a smooth curve. This curve, running through the beginning and end of the tunnel, is the first phase coarse path line.

[0055] For the second coarse path line, the staff opened the tunnel point cloud for the second phase, switched the view to a top-down view, and used the "polyline drawing tool" in the point cloud processing software to click the point line at intervals (e.g., 10 or 20 meters) along the center of the first lane of the tunnel. Finally, the point cloud processing software automatically connected these points with a smooth curve. This curve, running through the beginning and end of the tunnel, is the second phase coarse path line.

[0056] The node spacing of the coarse path line can be adjusted according to the tunnel orientation to avoid excessive spacing that could lead to fitting errors later. If an SHP vector file is imported to construct the coarse path line, the vector data can be edited to ensure consistency with the tunnel orientation.

[0057] In this embodiment, calculating the first phase distance along the path based on the first coarse path specifically includes projecting the first phase control point onto the first phase coarse path and calculating the length of the projection point from the path start point to obtain the first phase distance along the path of the first phase control point.

[0058] Specifically, taking a single first-phase control point (let's say control point A1) as an example, by vertically projecting control point A1 onto the coarse path line of the first phase, the point cloud processing software will automatically calculate its distance along the tunnel starting point, thus obtaining the first-phase distance along the tunnel for control point A1. The same process is applied to other first-phase control points, ultimately yielding the first-phase distance along the tunnel for each control point.

[0059] In this embodiment, calculating the second phase distance based on the second coarse path line specifically includes projecting the second phase control point onto the second phase coarse path line and calculating the length of the projection point from the path start point to obtain the second phase distance of the second phase control point.

[0060] Specifically, taking a second-phase control point (let's say control point B1) as an example, by vertically projecting control point B1 onto the coarse path line of the second phase, the point cloud processing software will automatically calculate its distance along the tunnel starting point, thus obtaining the second-phase distance along the tunnel for control point B1. The same process is applied to other second-phase control points, ultimately yielding the second-phase distance along the tunnel for each control point.

[0061] In this embodiment, the corrected distance along the line for the second-phase control points is calculated using the following formula:

[0062]

[0063] in, The corrected distance along the line for the second phase of control points. The distance along the second phase control point is the distance along the second phase line. The distance along the first phase of the control point is the distance along the first phase line.

[0064] Taking the 6th first-phase control point (let's say control point A6) and the 6th second-phase control point (let's say control point B6) as examples, if the distance along the line of control point A6 in the first phase is 75 meters and the distance along the line of control point B6 in the second phase is 74.85 meters, then the corrected distance along the line of control point B6 is 75 meters.

[0065] Once the corrected distances along the line for all the second-phase control points have been obtained, the tunnel point cloud of the second phase is moved according to the corrected distances along the line, thereby completing the coarse benchmark alignment of the two phase point clouds.

[0066] S3. Centerline Fitting: Generate tunnel cross-section, obtain centroid coordinates based on tunnel cross-section, and generate cross-section contour lines based on centroid coordinates; extract side vertical point clouds of maintenance lanes on both sides, obtain boundary lines of maintenance lanes on both sides based on side vertical point clouds of maintenance lanes on both sides, obtain local center Xf and spacing XL of boundary lines on both sides, filter out qualified tunnel cross-sections that do not exceed the first preset threshold based on spacing XL, and generate tunnel centerline based on centroid coordinates and local center Xf coordinates of qualified tunnel cross-sections.

[0067] In this embodiment, generating the tunnel cross section specifically includes generating the tunnel cross section based on the coarse path line and according to the preset cross section spacing and cross section width.

[0068] Taking the first phase of tunnel point cloud as an example, a cross section (with a width of 10 meters) is obtained by cutting along the first coarse path line at preset cross section intervals (e.g., 0.5 meters). The tunnel portion between two adjacent cross sections is a tunnel cross section (i.e., a tunnel segment). In addition, because the tunnel may have uphill or downhill slopes, the point cloud processing software does not cut perpendicular to the ground, but rather perpendicular to the tunnel's direction.

[0069] In this embodiment, obtaining the centroid coordinates based on the tunnel cross section specifically includes projecting the point cloud within the tunnel cross section onto a projection plane perpendicular to the coarse path line (which can be one of the cross sections of the tunnel segment), calling the DBSCAN algorithm to remove outliers within the projection plane, and calculating the centroid coordinates of the valid point cloud within the projection plane.

[0070] In this embodiment, generating the cross-sectional contour line based on the centroid coordinates specifically includes using an improved B-spline curve fitting algorithm and generating the cross-sectional contour line based on the centroid coordinates.

[0071] The default number of nodes for improving B-spline curve fitting is 50, and the default smoothness is 0.8, ranging from 0 to 1. More nodes and higher smoothness result in better fitting accuracy. Taking a tunnel cross-section as an example, the generated cross-sectional profile is as follows: Figure 4 As shown.

[0072] In this embodiment, extracting the vertical point cloud of the sides of the maintenance lanes specifically includes extracting the vertical point cloud of the sides of the maintenance lanes using a histogram algorithm.

[0073] In this embodiment, obtaining the boundary lines of the two maintenance lanes based on the side vertical point cloud of the two maintenance lanes specifically includes removing extreme points and then using the least squares algorithm to fit the boundary lines of the left and right maintenance lanes.

[0074] Taking a tunnel cross-section as an example, the vertical point cloud of the maintenance tunnel side is extracted using a histogram algorithm. Extreme points within the range of maximum and minimum elevation values ​​(0.2m) are removed. The left and right maintenance tunnel boundary lines are then fitted using a least squares algorithm. The local center Xf (i.e., the X-coordinate value of the midpoint between the two maintenance tunnel boundary lines) and the spacing XL (i.e., the width value of the two maintenance tunnel boundary lines) of the maintenance tunnel can be calculated using the left and right maintenance tunnel boundary lines.

[0075] In this embodiment, the first preset threshold is 0.5m. When the absolute value of the difference between the spacing XL of a certain tunnel section and the standard spacing is less than or equal to 0.5m, the tunnel section is a qualified tunnel section; otherwise, the tunnel section is not a qualified tunnel section.

[0076] For a qualified tunnel section, assuming its original centroid coordinates are "Xc, Yc, Zc", we now replace "Xc" with "Xf" to obtain the adjusted centroid coordinates "Xf, Yc, Zc". Each qualified tunnel section can obtain an adjusted centroid coordinate "Xf, Yc, Zc". Connecting all the adjusted centroid coordinates together yields the high-precision centerline of the tunnel, with the deviation between the centerline and the actual tunnel direction less than or equal to 0.2mm.

[0077] The tunnel centerline was obtained in the same way for the second phase of tunnel point cloud.

[0078] S4. Deformation Detection: Deformation detection of the tunnel lining, the tunnel pavement, and the tunnel maintenance road.

[0079] In this embodiment, the lining deformation detection of the tunnel cross section specifically includes:

[0080] Using the central axis of the first-phase tunnel as the center, the projection plane of the first-phase tunnel section is divided into several regions according to the angle, and the distance from the first-phase outline of each region to the center is obtained. Using the central axis of the second-phase tunnel as the center, the projection plane of the second-phase tunnel section is divided into several regions according to the angle, and the distance from the second-phase outline of each region to the center is obtained. The difference between distance one and distance two for the corresponding region is calculated as the lining deformation amount. It is determined whether the lining deformation amount exceeds the second preset threshold. When the lining deformation amount of a certain region exceeds the second preset threshold, the region is marked as lining abnormal.

[0081] Taking the No. 8 tunnel section of Phase I and Phase II tunnels as examples, firstly, the projection plane of the No. 8 tunnel section of Phase I is obtained (the projection plane already has the outline of the Phase I section). Using the central axis of Phase I tunnel as the center, the projection plane is divided into 120 regions at 3° intervals. The regions of Phase II correspond one-to-one with those of Phase I. For each region, the distance from the outline of Phase I to the center is obtained. Then, the projection plane of the No. 8 tunnel section of Phase II is obtained (the projection plane already has the outline of the Phase II section). Using the central axis of Phase II tunnel as the center, the projection plane is divided into 120 regions at 3° intervals. For each region, the distance from the outline of Phase II to the center is obtained. In this embodiment, the second preset threshold can be 7mm. Taking region 6 as an example, assuming the difference between the obtained distance 2 and distance 1 is 8.2mm, i.e., the lining deformation is 8.2mm, because the lining deformation of region 6 exceeds the second preset threshold, region 6 of the No. 8 tunnel section is determined to have an abnormal lining.

[0082] In this embodiment, deformation detection of the tunnel pavement specifically includes:

[0083] The first-phase tunnel is divided into segments, and the first-phase road surface elevation of each segment is calculated. The second-phase tunnel is also divided into segments, and the second-phase road surface elevation of each segment is calculated. The difference between the first-phase and second-phase road surface elevations of the corresponding segments is calculated as the road surface deformation. It is then determined whether the road surface deformation exceeds the third preset threshold. When the road surface deformation of a certain segment exceeds the third preset threshold, the segment is marked as an abnormal road surface.

[0084] Taking section 8 of the first-phase tunnel and section 8 of the second-phase tunnel as examples, assuming that section 8 of the first-phase tunnel has 10 road surface elevation monitoring points (point 1) and section 8 of the second-phase tunnel has 10 road surface elevation monitoring points (point 2), with a one-to-one correspondence between monitoring points 1 and 2. The difference between the first-phase and second-phase road surface elevations (i.e., the first road surface deformation) can be obtained through the first and second road surface elevation monitoring points; the difference between the second-phase and second-phase road surface elevations (i.e., the second road surface deformation) can be obtained through the second and second road surface elevation monitoring points (point 1 and point 2), and so on, until the difference between the tenth-phase and second-phase road surface elevations (point 1 and point 2) can be obtained (i.e., the tenth road surface deformation). When the elevation of the second-phase road surface is lower than that of the first-phase road surface, it indicates settlement; when the elevation of the second-phase road surface is higher than that of the first-phase road surface, it indicates uplift. The average and maximum road surface deformation can be obtained using 10 road surface deformation measurements. In this embodiment, the third preset threshold includes an average road surface deformation threshold and a maximum road surface deformation threshold. For example, the average road surface deformation threshold is 0.5 mm, and the maximum road surface deformation threshold is 1.5 mm. Assuming the average road surface deformation of section 8 is 0.3 mm and the maximum road surface deformation of section 8 is 1.2 mm, then because the average road surface deformation of section 8 is less than 0.5 mm and the maximum road surface deformation of section 8 is less than 1.5 mm, it is determined that there is no road surface abnormality in section 8.

[0085] In this embodiment, the inspection and deformation detection of the tunnel cross-section specifically includes:

[0086] Obtain the local center Xf and the highest elevation point of the first phase for each tunnel section of the first phase; obtain the local center Xf and the highest elevation point of the second phase for each tunnel section of the second phase; calculate the difference between the local center Xf of the first phase and the local center Xf of the second phase for the corresponding tunnel section as the deformation amount in the X direction of the maintenance road; calculate the difference between the highest elevation point of the first phase and the highest elevation point of the second phase for the corresponding tunnel section as the deformation amount in the Y direction of the maintenance road; determine whether the deformation amount in the X direction of the maintenance road exceeds the fourth preset threshold and whether the deformation amount in the Y direction of the maintenance road exceeds the fifth preset threshold; when the deformation amount in the X direction of the maintenance road of a certain tunnel section exceeds the fourth preset threshold or the deformation amount in the Y direction of the maintenance road exceeds the fifth preset threshold, mark the tunnel section as an abnormal maintenance road.

[0087] Taking the No. 6 tunnel section of the first-phase tunnel and the No. 6 tunnel section of the second-phase tunnel as examples, firstly, the local center Xf and the highest elevation point of the first-phase section of the No. 6 tunnel section of the first-phase tunnel are obtained. Then, the local center Xf and the highest elevation point of the second-phase section of the No. 6 tunnel section of the second-phase tunnel are obtained. Next, the difference between the local center Xf of the first-phase and the local center Xf of the second-phase tunnel is calculated as the X-direction deformation of the maintenance roadway of the No. 6 tunnel section. The difference between the highest elevation point of the first-phase and the highest elevation point of the second-phase tunnel section is calculated as the Y-direction deformation of the maintenance roadway. In this embodiment, the fourth preset threshold and the fifth preset threshold are both 0.5m. Assuming that the X-direction deformation of the maintenance roadway of the No. 6 tunnel section is 0.25m and the Y-direction deformation of the maintenance roadway of the No. 6 tunnel section is 0.18m, since the X-direction deformation of the maintenance roadway of the No. 6 tunnel section is less than 0.5m and the Y-direction deformation of the maintenance roadway of the No. 6 tunnel section is determined to have no maintenance roadway abnormality.

[0088] S5. Data Export: Export tunnel section information, deformation detection results, and centerline information.

[0089] In this embodiment, exporting tunnel cross-section information, deformation detection results, and centerline information specifically includes exporting these information in CAD, SHP vector, Excel report, and TXT analysis report formats, respectively. The exported Excel report includes a cross-section basic information table, a lining deformation table, a maintenance road parameter table, a pavement deformation table, and a comprehensive evaluation table. The TXT analysis report includes parameter settings, detection procedures, deformation data, and anomaly descriptions.

[0090] In addition, this embodiment also includes a visualization interaction module: used to display point cloud data, cross-sectional contours, central axis, and deformation detection results, providing view zooming, rotation, and translation operations, and supporting parameter setting and editing; the layout is clear, including a menu bar, function toolbar, data list area, visualization view area, and parameter setting area, making it easy to operate.

[0091] Figure 2 A block diagram of a tunnel LiDAR point cloud processing and deformation detection device according to an embodiment of the present invention is shown. The device includes:

[0092] The point cloud data preprocessing module is configured to load LiDAR point cloud data from the two phases of the tunnel, use the DBSCAN algorithm to automatically denoise the point cloud data, and manually retain the point cloud data of the target area through box selection denoising.

[0093] The benchmark correction module is configured to extract tunnel construction joints as control points, import control point data, construct coarse path lines, including a first coarse path line and a second coarse path line, calculate the first phase distance along the line based on the first coarse path line, calculate the second phase distance along the line based on the second coarse path line, and calculate the corrected distance along the line of the second phase control points based on the first phase distance and the second phase distance.

[0094] The central axis fitting module is configured to generate a tunnel cross section, obtain the centroid coordinates based on the tunnel cross section, generate the cross section outline based on the centroid coordinates, extract the side vertical point cloud of the maintenance lanes on both sides, obtain the boundary lines of the maintenance lanes on both sides based on the side vertical point cloud of the maintenance lanes on both sides, obtain the local center Xf and spacing XL of the boundary lines on both sides, filter out qualified tunnel cross sections that do not exceed the first preset threshold based on the spacing XL, and fit the centroid coordinates and local center Xf coordinates of the qualified tunnel cross sections to generate the tunnel central axis.

[0095] The deformation detection module is configured to detect the deformation of the tunnel lining, the tunnel pavement, and the tunnel maintenance road.

[0096] The data export module is configured to export tunnel cross-section information, deformation detection results, and centerline information.

[0097] Figure 3 A block diagram of an electronic device 300 according to some embodiments of the present invention is shown. The device 300 includes a processor 301, which performs various appropriate actions and processes based on computer program instructions loaded into random access memory (RAM) 303 according to computer program instructions stored in read-only memory (ROM) 302. Various programs and data required for the operation of the device 300 may also be stored in RAM 303. The processor 301, ROM 302, and RAM 303 are interconnected via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0098] The various processes and procedures described above, such as method 100, can be executed by processor 301. For example, in some embodiments, method 100 may be implemented as a software program tangibly contained in a machine-readable medium. In some embodiments, part or all of the software program may be loaded and / or installed on device 300 via ROM 302. When the software program is loaded into RAM 303 and executed by processor 301, one or more actions of method 100 described above may be performed.

[0099] The functions described above in this document can be performed at least in part by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: field programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload programmable logic devices (CPLDs), and so on.

[0100] The program code used to implement the methods of the present invention can be written in any combination of one or more programming languages. This program code can be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing device, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code can be executed entirely on the machine, partially on the machine, as a standalone software package partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0101] This invention can be a method, apparatus, system, and / or program product. The program product may include a machine-readable storage medium on which machine-readable program instructions for performing various aspects of the invention are loaded. The machine-readable program instructions described herein can be downloaded from the machine-readable storage medium to various computing / processing devices, or downloaded via a network, such as the Internet, a local area network, a wide area network, and / or a wireless network, to an external computer or external storage device. The network may include copper transmission cables, fiber optic transmissions, wireless transmissions, routers, firewalls, switches, gateway computers, and / or edge servers. A network adapter card or network interface in each computing / processing device receives the machine-readable program instructions from the network and forwards them for storage in the machine-readable storage medium of the respective computing / processing device.

[0102] Machine program instructions used to perform the operations of this invention may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk, C++, etc., and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The machine-readable program instructions may be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or may be connected to an external computer (e.g., via the Internet using an Internet service provider). In some embodiments, electronic circuitry, such as programmable logic circuitry, field-programmable gate arrays (FPGAs), or programmable logic arrays (PLAs), is personalized by utilizing state information from the machine-readable program instructions. This electronic circuitry can execute the machine-readable program instructions to implement various aspects of the invention.

[0103] In the context of this invention, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. Machine-readable media can include, but are not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing. Furthermore, although the operations are depicted in a specific order, this should be understood as requiring that such operations be performed in the specific order shown or in sequential order, or requiring that all illustrated operations be performed to achieve the desired result. In certain environments, multitasking and parallel processing may be advantageous. Similarly, while several specific implementation details are included in the above discussion, these should not be construed as limiting the scope of the invention. Certain features described in the context of individual embodiments may also be implemented in combination in a single embodiment. Conversely, various features described in the context of a single embodiment may also be implemented individually or in any suitable sub-combination in multiple embodiments.

[0104] Although the subject matter has been described using language specific to structural features and / or methodological logic, it should be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or actions described above. Rather, the specific features and actions described above are merely illustrative examples of implementing the claims.

Claims

1. A method for processing and detecting LiDAR point clouds in tunnels, characterized in that, include: S1. Point cloud data preprocessing: Load the LiDAR point cloud data of the two phases of the tunnel, use the DBSCAN algorithm to automatically denoise the point cloud data, and at the same time manually retain the point cloud data of the target area by selecting the area for denoising. S2. Benchmark Correction: Extract tunnel construction joints as control points and import control point data; construct coarse path lines, including a first coarse path line and a second coarse path line; calculate the first phase distance along the line based on the first coarse path line; calculate the second phase distance along the line based on the second coarse path line; calculate the corrected distance along the line for the second phase control points based on the first phase distance and the second phase distance; and complete the coarse benchmark alignment of the two phase point clouds based on the corrected distance. S3. Centerline Fitting: Generate tunnel cross-section, obtain centroid coordinates based on tunnel cross-section, and generate cross-section contour line based on centroid coordinates; extract side vertical point clouds of maintenance lanes on both sides, obtain boundary lines of maintenance lanes on both sides based on side vertical point clouds of maintenance lanes on both sides, obtain local center Xf and spacing XL of boundary lines on both sides, filter out qualified tunnel cross-sections that do not exceed the first preset threshold based on spacing XL, and generate tunnel centerline based on centroid coordinates and local center Xf coordinates of qualified tunnel cross-sections. S4. Deformation Detection: Deformation detection of the tunnel lining, the tunnel pavement, and the tunnel maintenance road. S5. Data Export: Export tunnel section information, deformation detection results, and centerline information.

2. The method according to claim 1, characterized in that, In step S1, the point cloud data is in LAS 1.2 / 1.3 format; automatic denoising of the point cloud data specifically involves setting the neighborhood radius and minimum number of neighborhood points to remove outliers and invalid points from the point cloud data; manual retention of the target area point cloud data through selection denoising specifically involves selecting the target area with the mouse to remove the point cloud data outside the selection.

3. The method according to claim 1, characterized in that, In step S2, importing control point data specifically includes importing control point data containing control point numbers, three-dimensional coordinates, and the phase to which they belong. The control point data includes first-phase control points and second-phase control points. Constructing coarse path lines specifically includes constructing first and second coarse path lines based on the path line of the first lane of the tunnel, either manually drawn or by importing SHP vector files. Calculating the first-phase distance along the path line based on the first coarse path line specifically includes projecting the first-phase control points onto the first-phase coarse path line and calculating the length of the projection point from the path start point to obtain the first-phase distance along the path line for the first-phase control points. Calculating the second-phase distance along the path line based on the second coarse path line specifically includes projecting the second-phase control points onto the second-phase coarse path line and calculating the length of the projection point from the path start point to obtain the second-phase distance along the path line for the second-phase control points.

4. The method according to claim 3, characterized in that, The corrected distance along the line for the second phase control points is calculated using the following formula: ， in, The corrected distance along the line for the second phase of control points. The distance along the second phase control point is the distance along the second phase line. The distance along the first phase of the control point is the distance along the first phase line.

5. The method according to claim 1, characterized in that, In step S3, generating the tunnel cross section specifically includes generating the tunnel cross section based on the coarse path line and according to the preset cross section spacing and width; obtaining the centroid coordinates based on the tunnel cross section specifically includes projecting the point cloud within the tunnel cross section onto a projection plane perpendicular to the coarse path line, calling the DBSCAN algorithm to remove outliers in the projection plane, and calculating the centroid coordinates of the effective point cloud in the projection plane; generating the cross section contour line based on the centroid coordinates specifically includes generating the cross section contour line using an improved B-spline curve fitting algorithm and the centroid coordinates as a reference; extracting the side vertical point cloud of the maintenance lanes on both sides specifically includes extracting the side vertical point cloud of the maintenance lanes using a histogram algorithm; obtaining the boundary lines of the maintenance lanes on both sides based on the side vertical point cloud of the maintenance lanes on both sides specifically includes removing extreme points and then using a least squares algorithm to fit the boundary lines of the left and right maintenance lanes.

6. The method according to claim 1, characterized in that, In step S4, the lining deformation detection of the tunnel section specifically includes: dividing the projection plane of the first-phase tunnel section into several regions according to angles, with the central axis of the first-phase tunnel as the center, and obtaining the distance from the first-phase outline of each region to the center of the circle; dividing the projection plane of the second-phase tunnel section into several regions according to angles, with the central axis of the second-phase tunnel as the center, and obtaining the distance from the second-phase outline of each region to the center of the circle; calculating the difference between the distance 1 and the distance 2 for the corresponding region as the lining deformation amount, determining whether the lining deformation amount exceeds a second preset threshold, and marking the region as lining abnormal when the lining deformation amount of a certain region exceeds the second preset threshold; The specific steps for detecting tunnel pavement deformation include: dividing the first-phase tunnel into segments and calculating the first-phase pavement elevation for each segment; dividing the second-phase tunnel into segments and calculating the second-phase pavement elevation for each segment; calculating the difference between the first-phase and second-phase pavement elevations for the corresponding segment as the pavement deformation amount; determining whether the pavement deformation amount exceeds a third preset threshold; and marking the pavement segment as abnormal when the pavement deformation amount of a certain segment exceeds the third preset threshold. The specific steps for detecting tunnel cross-section deformation include: obtaining the local center Xf and the highest elevation point of the first phase for each first-phase tunnel cross-section; obtaining the local center Xf and the highest elevation point of the second phase for each second-phase tunnel cross-section; calculating the difference between the local center Xf of the first phase and the local center Xf of the second phase for the corresponding tunnel cross-section as the deformation amount in the X direction of the inspection road; calculating the difference between the highest elevation point of the first phase and the highest elevation point of the second phase for the corresponding tunnel cross-section as the deformation amount in the Y direction of the inspection road; determining whether the deformation amount in the X direction of the inspection road exceeds a fourth preset threshold and whether the deformation amount in the Y direction of the inspection road exceeds a fifth preset threshold; and marking the tunnel cross-section as having an abnormal inspection road when the deformation amount in the X direction of the inspection road of a certain tunnel cross-section exceeds the fourth preset threshold or the deformation amount in the Y direction of the inspection road exceeds the fifth preset threshold.

7. The method according to claim 1, characterized in that, In step S5, exporting tunnel section information, deformation detection results, and centerline information specifically includes exporting the tunnel section information, deformation detection results, and centerline information in CAD, SHP vector, Excel report, and TXT analysis report formats, respectively.

8. A tunnel LiDAR point cloud processing and deformation detection device, characterized in that, include: The point cloud data preprocessing module is configured to load LiDAR point cloud data from the two phases of the tunnel, use the DBSCAN algorithm to automatically denoise the point cloud data, and manually retain the point cloud data of the target area through box selection denoising. The benchmark correction module is configured to extract tunnel construction joints as control points, import control point data, construct coarse path lines, which include a first coarse path line and a second coarse path line, calculate the first phase distance along the line based on the first coarse path line, calculate the second phase distance along the line based on the second coarse path line, and calculate the corrected distance along the line of the second phase control points based on the first phase distance and the second phase distance. The central axis fitting module is configured to generate a tunnel cross section, obtain the centroid coordinates based on the tunnel cross section, generate a cross section outline based on the centroid coordinates, extract the side vertical point clouds of the two maintenance lanes, obtain the boundary lines of the two maintenance lanes based on the side vertical point clouds of the two maintenance lanes, obtain the local center Xf and spacing XL of the boundary lines on both sides, filter out qualified tunnel cross sections that do not exceed a first preset threshold based on the spacing XL, and fit and generate the tunnel central axis based on the centroid coordinates and local center Xf coordinates of the qualified tunnel cross sections. The deformation detection module is configured to detect the deformation of the tunnel lining, the tunnel pavement, and the tunnel maintenance road. The data export module is configured to export tunnel cross-section information, deformation detection results, and centerline information.

9. A computer program product, characterized in that, Includes a computer program, which, when executed by a processor, implements the steps of the method according to any one of claims 1-7.

10. An electronic device, characterized in that, It includes one or more processors and memory associated with the one or more processors, the memory being used to store program instructions that, when read and executed by the one or more processors, perform the steps of the method according to any one of claims 1-7.

Images