Tunnel back break rapid analysis method and system based on model and point cloud matching

Abstract

The invention provides a tunnel over-break and under-break rapid analysis method and system based on model and point cloud matching, and relates to the technical field of tunnel engineering construction, and the method comprises the steps: obtaining tunnel point cloud data, and carrying out the preprocessing of the tunnel point cloud data, and obtaining high-quality actual measurement point cloud data; constructing a design model, and extracting points on each section contour line in the design model to obtain a design point cloud; according to the method, feature points of design point clouds and actual measurement point clouds are extracted, initial registration is carried out on the feature points of the design point clouds and the actual measurement point clouds, an improved ICP algorithm is utilized to optimize matching precision, in the matching process, matching errors are monitored in real time, error correction and dynamic optimization are carried out, and finally a matching result is output and obtained; and based on a matching result, performing spatial difference analysis on the aligned actually measured point cloud and the design model, and accurately calculating the back-break volume of each point or region, thereby realizing quantitative evaluation of the back-break condition of the tunnel. The quality control level of tunnel construction is effectively improved.

Description

Technical Field

[0001] This disclosure relates to the field of tunnel engineering construction technology, specifically to a rapid analysis method and system for tunnel over-excavation and under-excavation based on model and point cloud matching. Background Technology

[0002] The statements in this section are merely background information relating to this disclosure and do not necessarily constitute prior art.

[0003] In tunnel construction, over-excavation and under-excavation control are key factors affecting project quality, construction safety, and economic benefits. Over-excavation leads to increased support costs and extended construction periods, while under-excavation may require secondary treatment and even affect the stability of the tunnel structure. Traditional over-excavation and under-excavation detection methods mainly rely on single-point measurement equipment such as total stations and cross-section instruments, collecting limited feature point data and comparing it with the design cross-section for analysis. However, these methods suffer from drawbacks such as low measurement efficiency, insufficient data density, and high levels of manual intervention, making it difficult to meet the demands of modern tunnel engineering for high-precision, real-time, and automated detection.

[0004] In recent years, 3D laser scanning technology has been gradually introduced into the field of tunnel engineering inspection due to its advantages such as high precision, high efficiency, and non-contact measurement. This technology can quickly acquire massive amounts of point cloud data from the tunnel surface, providing more comprehensive spatial information for over-excavation and under-excavation analysis. However, existing laser scanning-based methods for calculating over-excavation and under-excavation still have some limitations: (I) On the one hand, point cloud data processing is inefficient and subject to significant noise interference. The amount of tunnel point cloud data acquired by 3D laser scanning is enormous, and it is often affected by the construction environment (such as dust, lighting, and equipment vibration), containing a large amount of noise and redundant data. Traditional data filtering and denoising algorithms have high computational complexity, making it difficult to achieve rapid processing while maintaining accuracy. This results in low efficiency in over-excavation and under-excavation analysis, failing to meet the needs of real-time construction monitoring. How to efficiently remove noise, simplify point cloud data, and retain effective tunnel contour features is a key technical challenge for achieving rapid calculation of over-excavation and under-excavation.

[0005] (ii) On the other hand, there is the problem of automated and accurate matching between the design model and the measured point cloud. Existing over-excavation and under-excavation analysis methods usually rely on manual selection of reference surfaces or segmental comparison, lacking an automated global matching strategy, resulting in strong subjectivity in the calculation process and significant error accumulation. Since local deformation or construction deviations may occur during tunnel excavation, how to achieve automatic alignment between the measured point cloud and the design model, and accurately quantify the over-excavation and under-excavation volume, is a technical challenge that has not yet been effectively solved.

[0006] (III) In addition, how to dynamically optimize the matching algorithm to adapt to the geometric characteristics of different tunnel sections is also an important challenge to improve the calculation accuracy. Existing methods mostly rely on manual selection of reference surfaces or local comparisons, and lack automated global over-excavation and under-excavation assessment methods, resulting in insufficient reliability and consistency of calculation results. Summary of the Invention

[0007] To address the aforementioned issues, this disclosure proposes a rapid analysis method and system for tunnel over-excavation and under-excavation based on model-point cloud matching. By optimizing the point cloud processing algorithm and combining it with an improved Iterative Closest Point (ICP) algorithm, the design model and the measured point cloud are automatically aligned, achieving automated matching and calculation, improving detection efficiency and accuracy, and providing reliable technical support for tunnel construction quality control.

[0008] According to some embodiments, the present disclosure adopts the following technical solutions: A rapid analysis method for tunnel over-excavation and under-excavation based on model and point cloud matching includes: The tunnel point cloud data is acquired and preprocessed to obtain high-quality measured point cloud data. Construct a design model, extract points from the contour lines of each section in the design model, and obtain the design point cloud; Feature points are extracted from the design point cloud and the measured point cloud, and spatial index structures for the feature points are established respectively. Based on the spatial indexing results, the feature points of the design point cloud and the measured point cloud are initially registered, and the improved ICP algorithm is used to optimize the matching accuracy. The closest corresponding point between the measured point cloud and the design model is iteratively found, and the distance error between them is minimized. During the matching process, the matching error is monitored in real time, and error correction and dynamic optimization are performed. Finally, the matching result is output. Based on the matching results, spatial difference analysis is performed between the aligned measured point cloud and the design model to accurately calculate the over- or under-excavation volume of each point or region, thereby achieving a quantitative assessment of the over- or under-excavation situation of the tunnel.

[0009] According to some embodiments, the present disclosure adopts the following technical solutions: A rapid analysis system for tunnel over- and under-excavation based on model and point cloud matching includes: The point cloud acquisition module is used to acquire tunnel point cloud data and preprocess it to obtain high-quality measured point cloud data. The feature point extraction module is used to construct the design model, extract points on the contour lines of each section in the design model to obtain the design point cloud; extract feature points from the design point cloud and the measured point cloud, and establish spatial index structures for the feature points respectively. The registration module is used to perform initial registration of feature points between the design point cloud and the measured point cloud based on the spatial index results. It also uses an improved ICP algorithm to optimize the matching accuracy, iteratively finds the nearest corresponding point between the measured point cloud and the design model, and minimizes the distance error between them. During the matching process, the matching error is monitored in real time, and error correction and dynamic optimization are performed. Finally, the matching result is output. The over- and under-excavation volume quantification module is used to perform spatial difference analysis between the aligned measured point cloud and the design model based on the matching results, accurately calculate the over- and under-excavation volume of each point or region, and thus realize the quantitative assessment of the over- and under-excavation situation of the tunnel.

[0010] According to some embodiments, the present disclosure adopts the following technical solutions: A computer program product includes a computer program that, when executed by a processor, implements the aforementioned method for rapid analysis of tunnel over-excavation and under-excavation based on model-point cloud matching.

[0011] According to some embodiments, the present disclosure adopts the following technical solutions: A non-transitory computer-readable storage medium is provided for storing computer instructions, which, when executed by a processor, implement the aforementioned method for rapid analysis of tunnel over-excavation and under-excavation based on model-point cloud matching.

[0012] According to some embodiments, the present disclosure adopts the following technical solutions: An electronic device includes a processor, a memory, and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory to enable the electronic device to perform the aforementioned method for rapid analysis of tunnel over-excavation and under-excavation based on model and point cloud matching.

[0013] Compared with the prior art, the beneficial effects of this disclosure are as follows: This disclosed method for rapid analysis of tunnel over- and under-excavation based on model-point cloud matching achieves efficient and accurate model-point cloud alignment through feature extraction, ICP and its improved algorithm, error correction, and dynamic optimization. This provides a solid data foundation and reliable technical support for tunnel over- and under-excavation calculations. The method has wide applicability and practicality in tunnel cross-sections of different shapes and complex construction environments, effectively improving the quality control level and economic benefits of tunnel construction.

[0014] This disclosure presents a rapid analysis method for tunnel over- and under-excavation based on model and point cloud matching. To ensure the accuracy and efficiency of over- and under-excavation volume calculation, a series of data processing and optimization steps are performed. First, filtering and denoising are conducted using an adaptive filtering algorithm to remove noise points from the point cloud data while retaining key features of the tunnel outline. Then, voxel mesh downsampling technology is used to simplify the point cloud data into representative feature points, reducing data volume and improving computational efficiency. Next, feature extraction and matching are performed, extracting key geometric feature points, such as contour lines and turning points, from the design model and measured point cloud. The Iterative Closest Point (ICP) algorithm and its improved version are used for automatic alignment to ensure matching accuracy and efficiency. Finally, error correction and dynamic optimization are implemented, introducing an error feedback mechanism during the matching process to monitor and correct matching errors in real time. Based on different tunnel cross-sectional geometric features (such as circular, horseshoe-shaped, etc.), the parameters and strategies of the matching algorithm are dynamically adjusted to optimize the feature extraction and matching process.

[0015] This disclosure presents a rapid analysis method for tunnel over- and under-excavation based on model and point cloud matching. The calculation results are output in a structured data format, including information such as the over- and under-excavation volume of each unit, the total volume, and the distribution of over-excavation and under-excavation areas. The calculation results are fed back to a visualization platform in real time, and the over- and under-excavation trends are analyzed. Combined with construction progress and historical data, the changing trends of over- and under-excavation are dynamically analyzed to predict potential construction problems and provide timely decision support for construction personnel. Attached Figure Description

[0016] The accompanying drawings, which form part of this disclosure, are used to provide a further understanding of this disclosure. The illustrative embodiments of this disclosure and their descriptions are used to explain this disclosure and do not constitute an undue limitation of this disclosure.

[0017] Figure 1 This is a flowchart of a rapid analysis method for tunnel over-excavation and under-excavation based on model and point cloud matching, according to an embodiment of this disclosure. Detailed Implementation

[0018] The present disclosure will be further described below with reference to the accompanying drawings and embodiments.

[0019] It should be noted that the following detailed descriptions are illustrative and intended to provide further explanation of this disclosure. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure pertains.

[0020] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the exemplary embodiments according to this disclosure. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms “comprising” and / or “including” are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0021] Example 1 One embodiment of this disclosure provides a rapid analysis method for tunnel over-excavation and under-excavation based on model-point cloud matching. The method steps are as follows: Step 1: Acquire tunnel point cloud data and preprocess it to obtain high-quality measured point cloud data; Step 2: Construct a design model and extract points from the contour lines of each section in the design model to obtain the design point cloud; Step 3: Extract feature points from the design point cloud and the measured point cloud, and establish spatial index structures for the feature points respectively; Step 4: Based on the spatial indexing results, perform initial registration of feature points between the design point cloud and the measured point cloud, and use the improved ICP algorithm to optimize the matching accuracy. Iteratively find the nearest corresponding point between the measured point cloud and the design model, and minimize the distance error between them. During the matching process, monitor the matching error in real time, perform error correction and dynamic optimization, and finally output the matching result. Step 5: Based on the matching results, perform spatial difference analysis between the aligned measured point cloud and the design model to accurately calculate the over- or under-excavation volume of each point or region, thereby achieving a quantitative assessment of the tunnel's over- or under-excavation situation.

[0022] As one embodiment, the rapid analysis method for tunnel over-excavation and under-excavation based on model and point cloud matching disclosed herein improves detection efficiency and accuracy by optimizing the point cloud processing algorithm and realizing automated matching and calculation, providing reliable technical support for tunnel construction quality control. The specific implementation process is as follows: Step 1: Acquire tunnel point cloud data and preprocess it to obtain high-quality measured point cloud data, specifically including the following: Step 11: A 3D laser scanner acquires tunnel point cloud data; First, a 3D laser scanner is set up with multiple stations inside the tunnel, each collecting point cloud data within a defined range. The 3D laser scanner measures distance by emitting a laser beam and receiving reflected light, thereby obtaining point cloud information such as the 3D coordinates (X, Y, Z) and reflection intensity of points on the tunnel surface.

[0023] Step 12: The acquired raw point cloud data contains a large amount of noise and redundant data, requiring preprocessing. The first step in preprocessing is noise removal, including: An adaptive filtering algorithm is used to process point cloud data. The neighborhood of a data point is defined as having a radius of centered at the current point. r The set of all points within the neighborhood of the sphere is represented as:

[0024] in, For the current point, For points within the neighborhood, Let be the Euclidean distance between the two points.

[0025] Furthermore, for each data point Calculate the local mean and standard deviation of points in its neighborhood:

[0026]

[0027] in, This represents the number of points within the neighborhood.

[0028] Based on the local mean and standard deviation, set an appropriate threshold. The deviation from the mean will exceed Points that are identified as noise points are removed:

[0029] This disclosure effectively removes noise points while preserving the key features of the tunnel outline through the aforementioned filtering method.

[0030] Step 13: The second preprocessing step is to simplify the denoised point cloud data, including: A voxel downsampling technique is used to divide the 3D space into a voxel grid. Each grid cell uses a center point or feature point to represent the point cloud information within that region. The size of the voxel grid can be adjusted according to the data density and the required accuracy.

[0031] For each voxel grid, calculate the average coordinates of all points within the grid as the coordinates of the feature point:

[0032]

[0033]

[0034] in, n This represents the number of points within the voxel grid. , , These are the coordinates of each point.

[0035] This disclosure utilizes voxel downsampling to significantly reduce the amount of data while preserving the basic shape and key features of the tunnel profile.

[0036] In addition, data quality assessment can be considered during data acquisition and preprocessing. Indicators such as point cloud density, noise level, and contour fidelity can be used to evaluate data quality. Point cloud density is measured by calculating the number of points per unit area, noise level is assessed by calculating the standard deviation of the point cloud data, and contour fidelity is determined by comparing the original data with the preprocessed data.

[0037] Step 2: Construct the design model and extract points from the contour lines of each section in the design model to obtain the design point cloud. Specific details include the following: Step 21: Build the design model and prepare the data.

[0038] The design model exists in the form of a 3D CAD model, containing information such as the tunnel's cross-sectional contours, axes, and key geometric feature points. The design model is then converted into a discrete point cloud or mesh format and processed uniformly with the measured point cloud data.

[0039] Step 22: Extract points from the profile lines of each section in the design model to obtain the design point cloud, including: For each cross-sectional profile in the CAD model, discrete sampling is performed at set arc length intervals to extract discrete points on the profile line, forming a design point cloud dataset, where each point... Includes coordinate information .

[0040] The discrete sampling is performed using an arc-length equidistant sampling method, and the sampling interval Δs is determined based on the tunnel cross-section dimensions and the required matching accuracy. Preferably, the sampling interval Δs is 0.01m to 0.20m; more preferably, it is 0.03m to 0.10m; and in conventional railway or highway tunnel cross-sections, it is preferably 0.05m.

[0041] In one embodiment, the sampling interval Δs is adaptively determined based on the cross-sectional perimeter L: Δs = L / N, where N is the number of sampling points on the cross-section, preferably 200–1500, more preferably 300–800. This method ensures that tunnel cross-sections of different sizes have a relatively consistent sampling density.

[0042] The design point cloud data is standardized by shifting the origin of the coordinate system to the starting point of the tunnel or a key feature point, thus unifying the scale of the coordinate system.

[0043] The measured point cloud data is obtained from step 1. After preliminary filtering, denoising, and simplification, a high-quality measured point cloud dataset is obtained. , where each point It also contains coordinate information. .

[0044] Furthermore, coordinate system correction is performed on the measured point cloud data to ensure that it is consistent with the coordinate system of the design model.

[0045] Step 3: Extract feature points from the design point cloud and the measured point cloud, and establish spatial index structures for the feature points respectively. Specific details include: Step 31: Extract feature points from the design point cloud and the measured point cloud, including: First, contour lines are extracted. For both the design model and the measured point cloud, boundary detection algorithms are used to extract the contour lines of the tunnel cross-section.

[0046] Furthermore, for the design model, points on the theoretical contour line can be directly extracted based on its parametric representation; For measured point clouds, geometric features such as local curvature and normal vectors are used to identify contour lines. For example, in a measured point cloud, the normal vector of each point is calculated. Points on the contour line are determined by analyzing the changes in the direction of the normal vector. Points on the contour line satisfy a certain threshold for the change in the direction of the normal vector, i.e., the angle between the normal vectors of adjacent points. Greater than the set threshold :

[0047] Furthermore, turning points are extracted. On the tunnel outline, turning points are points where the outline direction changes significantly.

[0048] For design models, inflection points are usually predefined by the designer or extracted according to geometric rules; For the measured point cloud, a curvature-based inflection point detection algorithm is used. The curvature of each point on the contour line is calculated. Points with greater curvature are considered inflection points. A curvature threshold is set. ,when > When that point is reached, mark it as a turning point.

[0049] Step 32: Establish the spatial index structure for each feature point, including: After feature point extraction from the design model and measured point cloud is completed, a KD-Tree spatial index structure is constructed for the design feature point set to improve the efficiency of subsequent nearest neighbor search.

[0050] Let the feature point set of the design model be:

[0051] The feature point set of the measured point cloud is as follows:

[0052] Any feature point is represented as:

[0053] A three-dimensional KD-Tree index structure is constructed using the aforementioned feature point sets as input. The KD-Tree is a recursively partitioned binary tree structure. Its construction process includes: selecting the dimension with the largest coordinate variance in the current point set as the partition dimension, sorting the point set according to this dimension, selecting the median as the current node, and partitioning the remaining points into the left and right subsets. The process of recursively constructing child nodes continues until the termination condition is met.

[0054] During the matching process, for any query point Solve using a KD-Tree nearest neighbor search:

[0055] Where the Euclidean distance is ; The KD-Tree spatial index structure described above can significantly improve query efficiency in the feature point matching stage.

[0056] Step 4: Based on the spatial indexing results, perform initial registration of feature points between the design point cloud and the measured point cloud. Utilize an improved ICP algorithm to optimize matching accuracy, iteratively finding the closest corresponding points between the measured point cloud and the design model, and minimizing the distance error between them. During the matching process, monitor the matching error in real time, perform error correction and dynamic optimization, and finally output the matching result, as detailed below: Step 41: Based on the spatial indexing results, perform initial registration of feature points between the design point cloud and the measured point cloud, and optimize the matching accuracy using the improved ICP algorithm. Iteratively find the nearest corresponding point between the measured point cloud and the design model, and minimize the distance error between them, including: First, initial registration is performed based on the geometric relationships between feature points. Several pairs of matching feature points are selected from the feature point sets of the design model and the measured point cloud, and the initial spatial transformation matrix is ​​calculated using these matching point pairs. Including rotation matrix Translation vector This initial transformation matrix roughly aligns the measured point cloud to the coordinate system of the design model.

[0057] Furthermore, the Iterative Closest Point (ICP) algorithm is used for optimization. Based on the initial registration, the ICP algorithm is applied to further optimize the matching accuracy. The core idea of ​​the ICP algorithm is to iteratively find the closest corresponding points between the measured point cloud and the design model, and minimize the distance error between them. The specific steps are as follows: For the measured point cloud in the current iteration step In designing point cloud models For each point Find the nearest neighbor Nearest neighbor search can be quickly performed using a pre-built spatial index structure.

[0058] Calculate the sum of squared distance errors between the measured point cloud and the design model:

[0059] Calculate the new rotation matrix based on the corresponding point pairs. Translation vector This causes the error Minimize. This can be solved using singular value decomposition (SVD) or other optimization methods.

[0060] Update the coordinates of the measured point cloud:

[0061] Repeat the above steps until the error is found. Below the set threshold Or reach the maximum number of iterations .

[0062] This disclosure improves the above-mentioned ICP algorithm by optimizing the matching process using the improved ICP algorithm, which includes the following steps: In each iteration, the nearest neighbor search is first performed on the set of feature points in the measured point cloud to establish a set of feature-corresponding point pairs. Based on this, corresponding point matching is then performed on the remaining points, and abnormal corresponding points with a distance exceeding a preset threshold are removed. Let the error function for the k-th iteration be:

[0063] The Euclidean distance is:

[0064] The updated rotation matrix is ​​computed using Singular Value Decomposition (SVD) based on the set of corresponding point pairs. Translation vector And update the point cloud coordinates:

[0065] Simultaneously, a hierarchical matching strategy is adopted, which performs preliminary registration under coarse resolution point cloud and then fine optimization under fine resolution point cloud until the error meets the threshold condition or reaches the maximum number of iterations.

[0066] Step 42: During the matching process, the matching error is monitored in real time, and error correction and dynamic optimization are performed. The final output is the matching result, including: An error feedback mechanism monitors matching errors in real time during the matching process, including rotation error, translation error, and overall distance error. The error calculated in the next iteration is then fed back. Analyze its changing trend. If the error increases instead of decreasing in several consecutive iterations, or exceeds a certain reasonable range, trigger the error feedback mechanism.

[0067] Based on the error feedback, adjust the parameters of the matching algorithm, such as adjusting the search range of corresponding points in the ICP algorithm, increasing or decreasing the weight of feature points, etc.

[0068] Furthermore, this disclosure proposes a dynamic optimization strategy for different cross-sectional shapes. For circular tunnel cross-sections, optimization is performed utilizing their axisymmetric properties. During the feature extraction stage, in addition to extracting the contour lines and inflection points, the center and radius of the cross-section are calculated. During the matching process, the center of the measured cross-section is aligned with the center of the designed cross-section, while adjusting for radius differences. The matching results are optimized by minimizing center offset and radius differences. (Center offset error) and radius difference error They can be represented as follows:

[0069]

[0070] in, and These are the coordinates of the center of the circle for the measured and designed cross sections, respectively. and These are the measured and designed radii, respectively.

[0071] For horseshoe-shaped cross-sections, which have unique geometry, the top and bottom are wider, while the sides are narrower. During the matching process, the matching accuracy of the top and bottom contours is of paramount importance.

[0072] A region-based matching strategy is adopted, dividing the cross-section into top, bottom, and side regions for feature extraction and matching. The matching results of each region are then combined for overall optimization. For the top region, due to its complex contour changes, the feature point extraction density is increased, and more refined ICP iteration parameters are used. For the side regions, their relatively simple geometry is utilized to quickly perform matching and constrain the overall matching direction.

[0073] Specifically, for the top region, due to the large curvature change at the arch and the significant impact of construction disturbance, the feature point extraction density is first increased and the sampling interval is reduced. During the matching process, a smaller distance threshold and convergence threshold are set, and the maximum number of ICP iterations is increased to improve the matching accuracy in this region.

[0074] For the two side regions, since the geometry is relatively smooth and stable, a conventional sampling density is used for feature extraction, and a large matching distance threshold and a small number of iterations are set to achieve fast matching. At the same time, the overall orientation of the side wall region is used to constrain the rotation direction to reduce the overall attitude shift.

[0075] For the bottom region, the focus is on ensuring vertical alignment accuracy. During the matching process, the vertical offset is constrained to stabilize the registration results in the overall height direction.

[0076] After local matching is completed in each region, the region transformation parameters are obtained respectively. The transformation results of each region are then weighted and fused to calculate the overall optimal rotation matrix and translation vector, which are used as the global update parameters for the current iteration stage, thereby achieving a combination of regional matching and overall optimization.

[0077] This disclosure completes the automated and accurate matching of the design model and the measured point cloud, and outputs the matching results, including the final spatial transformation matrix. Matching error reports (such as error distribution in each region, overall error indicators, etc.) and joint visualization of aligned measured point cloud data and design model.

[0078] The matching results will provide an accurate basis for subsequent over- and under-excavation calculations. By performing spatial difference analysis between the aligned measured point cloud and the design model, the over- and under-excavation volume of each point or region can be accurately calculated, thereby achieving a quantitative assessment of the tunnel's over- and under-excavation situation. Simultaneously, the matching results can also be fed back into the construction process, guiding construction personnel to adjust construction parameters and methods in a timely manner and optimize tunnel construction quality control.

[0079] The automated and precise matching scheme between the design model and the measured point cloud achieves efficient and accurate alignment of the model and the point cloud through feature extraction, ICP and its improved algorithm, error correction, and dynamic optimization. This provides a solid data foundation and reliable technical support for tunnel over-excavation and under-excavation calculations. This scheme has wide applicability and practicality in tunnel cross-sections of different shapes and complex construction environments, effectively improving the quality control level and economic benefits of tunnel construction.

[0080] Step 5: Based on the matching results, perform spatial difference analysis between the aligned measured point cloud and the design model to accurately calculate the over- or under-excavation volume of each point or region, thereby achieving a quantitative assessment of the tunnel's over- or under-excavation situation.

[0081] Specifically, the core of the spatial interpolation analysis algorithm lies in accurately calculating the spatial difference between the design model and the measured point cloud. By constructing a voxel model, the complex three-dimensional spatial problem is decomposed into multiple computable unit volumes.

[0082] Spatial difference calculation: After matching the design model with the measured point cloud, it is necessary to calculate the spatial difference between each measured point and the design model. This difference is defined as the vertical distance from the measured point to the design model, and can be expressed as:

[0083] in, The first point in the measured point cloud One point, Represents the surface of the design model. This represents the perpendicular distance from a point to a surface.

[0084] As one implementation, to accurately calculate the over- and under-excavation volumes, triangulation meshes are constructed from both the design model and the measured point cloud. The triangulation meshes are constructed based on the Delaunay triangulation algorithm, ensuring that the vertices of each triangle are distributed as uniformly as possible in space. For the design model, the vertices of the triangulation mesh are the design points. For the measured point cloud, the vertices of the triangular network are the measured points. .

[0085] In the process of constructing a triangulated network, the area and normal vector of each triangle can be expressed as:

[0086]

[0087] in, , , These are the three vertices of the triangle. Let the area be the triangle. Let be the normal vector of the triangle.

[0088] As one example, a voxel model is constructed. A voxel model is a three-dimensional discrete representation method based on voxels (volume pixels). The space occupied by the design model and the measured point cloud is divided into a regular voxel grid, and the size of each voxel is determined according to the required accuracy and computational efficiency. The construction process of the voxel model is as follows: (1) Determine the size of the voxel grid , , .

[0089] (2) Divide the space into a voxel grid, with the coordinates of each voxel being... ,in , , It is an integer.

[0090] (3) For each point in the design model and the measured point cloud, determine the voxel to which it belongs and mark the voxel as the containing point.

[0091] (4) Calculate the number of points and average coordinates within each voxel for subsequent volume calculation.

[0092] Furthermore, the volume of each element is calculated. For each triangle or voxel, its over- or under-excavation volume is calculated. For a triangular element, the over- or under-excavation volume can be expressed as:

[0093] in, Let the area be the triangle. This is the vertical distance from the center point of the triangle to the design model.

[0094] For a voxel unit, the over- or under-dug volume can be expressed as:

[0095] in, Let V be the volume of a voxel. and These represent the distance values ​​at the voxel location, as measured in the point cloud and in the design model, respectively.

[0096] Furthermore, the over- and under-excavation volume calculation and output involves calculating the over- and under-excavation volume for each triangular or voxel unit and summing them to obtain the total over- and under-excavation volume of the entire tunnel cross-section.

[0097] in, The total number of triangular unit cells or voxel unit cells. For the first The over- or under-excavation volume of each unit.

[0098] Finally, the results are output and visualized. The over-excavation and under-excavation calculation results are output in a structured data format, including information such as the over-excavation and under-excavation volume of each unit, the total volume, and the distribution of over-excavation and under-excavation areas. The over-excavation and under-excavation calculation results are fed back to the visualization platform in real time, displaying the distribution of over-excavation and under-excavation areas of the tunnel cross-section in an intuitive color-coded or 3D graphic format. Over-excavation areas are represented in red, under-excavation areas in blue, and normal areas in green. Furthermore, the over-excavation and under-excavation trends are analyzed. Combining construction progress and historical data, the changing trends of over-excavation and under-excavation are dynamically analyzed to predict potential construction problems and provide timely decision support for construction personnel.

[0099] Example 2 One embodiment of this disclosure provides a rapid analysis system for tunnel over-excavation and under-excavation based on model and point cloud matching, including: The point cloud acquisition module is used to acquire tunnel point cloud data and preprocess it to obtain high-quality measured point cloud data. The feature point extraction module is used to construct the design model, extract points on the contour lines of each section in the design model to obtain the design point cloud; extract feature points from the design point cloud and the measured point cloud, and establish spatial index structures for the feature points respectively. The registration module is used to perform initial registration of feature points between the design point cloud and the measured point cloud based on the spatial index results. It also uses an improved ICP algorithm to optimize the matching accuracy, iteratively finds the nearest corresponding point between the measured point cloud and the design model, and minimizes the distance error between them. During the matching process, the matching error is monitored in real time, and error correction and dynamic optimization are performed. Finally, the matching result is output. The over- and under-excavation volume quantification module is used to perform spatial difference analysis between the aligned measured point cloud and the design model based on the matching results, accurately calculate the over- and under-excavation volume of each point or region, and thus realize the quantitative assessment of the over- and under-excavation situation of the tunnel.

[0100] Example 3 One embodiment of this disclosure provides a computer program product, including a computer program that, when executed by a processor, implements the aforementioned method for rapid analysis of tunnel over-excavation and under-excavation based on model-point cloud matching.

[0101] Example 4 One embodiment of this disclosure provides a non-transitory computer-readable storage medium for storing computer instructions. When these computer instructions are executed by a processor, they implement the aforementioned method for rapid analysis of tunnel over-excavation and under-excavation based on model-point cloud matching.

[0102] Example 5 One embodiment of this disclosure provides an electronic device, including a processor, a memory, and a computer program; wherein the processor is connected to the memory, and the computer program is stored in the memory. When the electronic device is running, the processor executes the computer program stored in the memory to enable the electronic device to perform the aforementioned method for rapid analysis of tunnel over-excavation and under-excavation based on model and point cloud matching.

[0103] This disclosure is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this disclosure. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create a machine for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0104] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0105] While the specific embodiments of this disclosure have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of this disclosure. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of this disclosure are still within the scope of protection of this disclosure.

Claims

1. A rapid analysis method for tunnel over-excavation and under-excavation based on model and point cloud matching, characterized in that, include: The tunnel point cloud data is acquired and preprocessed to obtain high-quality measured point cloud data. Construct a design model, extract points from the contour lines of each section in the design model, and obtain the design point cloud; Feature points are extracted from the design point cloud and the measured point cloud, and spatial index structures for the feature points are established respectively. Based on the spatial indexing results, the feature points of the design point cloud and the measured point cloud are initially registered, and the improved ICP algorithm is used to optimize the matching accuracy. The closest corresponding point between the measured point cloud and the design model is iteratively found, and the distance error between them is minimized. During the matching process, the matching error is monitored in real time, and error correction and dynamic optimization are performed. Finally, the matching result is output. Based on the matching results, spatial difference analysis is performed between the aligned measured point cloud and the design model to accurately calculate the over- or under-excavation volume of each point or region, thereby achieving a quantitative assessment of the over- or under-excavation situation of the tunnel.

2. The rapid analysis method for tunnel over-excavation and under-excavation based on model and point cloud matching as described in claim 1, characterized in that, The process of acquiring tunnel point cloud data and preprocessing it to obtain high-quality measured point cloud data includes: The 3D laser scanner sets up multiple stations inside the tunnel. Each station collects point cloud data within a set range. The 3D laser scanner measures distance by emitting laser beams and receiving reflected light, thereby obtaining the 3D coordinates and reflection intensity of points on the tunnel surface. An adaptive filtering algorithm is used to remove noise points from the point cloud data, and voxel mesh downsampling technology is used to simplify the denoised point cloud data to obtain high-quality measured point cloud data.

3. The rapid analysis method for tunnel over-excavation and under-excavation based on model and point cloud matching as described in claim 1, characterized in that, The construction of the design model, extracting points from the contour lines of each section in the design model to obtain the design point cloud, includes: The design model is constructed in the form of a 3D CAD model, including the cross-sectional contours, axes, and key geometric feature points of the tunnel. The design model is converted into a discrete point cloud or mesh format. For each cross-sectional contour in the design model, points on the contour line are extracted at set intervals to form a design point cloud dataset. The design point cloud data is standardized to unify the coordinate system scale.

4. The rapid analysis method for tunnel over-excavation and under-excavation based on model and point cloud matching as described in claim 1, characterized in that, The extraction of feature points from the design point cloud and the measured point cloud, and the establishment of spatial index structures for the feature points, includes: For both the design model and the measured point cloud, boundary detection algorithms are used to extract the contour lines of the tunnel cross-section. For the design model, feature points on the theoretical contour line are directly extracted based on its parametric representation; For measured point clouds, feature points of the contour lines are identified using local curvature and normal vector geometric features. After the feature points of the design model and the measured point cloud are extracted, a spatial index structure for the feature points is established.

5. The rapid analysis method for tunnel over-excavation and under-excavation based on model and point cloud matching as described in claim 1, characterized in that, Based on the spatial indexing results, the feature points of the design point cloud and the measured point cloud are initially registered, and the matching accuracy is optimized using an improved ICP algorithm. The process iteratively finds the nearest corresponding point between the measured point cloud and the design model, and minimizes the distance error between them, including: Initial registration is performed based on the geometric relationship between feature points. Several pairs of matching feature points are selected from the feature point sets of the design model and the measured point cloud, and the initial spatial transformation matrix, including the rotation matrix and translation vector, is calculated using the matching point pairs. The initial transformation matrix aligns the measured point cloud to the coordinate system of the design model; Based on the initial registration, the improved ICP algorithm is applied to optimize the matching accuracy. In each iteration, the nearest neighbor search is first performed on the set of feature points in the measured point cloud to establish a set of feature corresponding point pairs. On this basis, the corresponding point matching is performed on the remaining points, and abnormal corresponding points with a distance exceeding a preset threshold are removed.

6. The rapid analysis method for tunnel over-excavation and under-excavation based on model and point cloud matching as described in claim 1, characterized in that, Based on the matching results, spatial difference analysis is performed between the aligned measured point cloud and the design model to accurately calculate the over- or under-excavation volume of each point or region, thereby achieving a quantitative assessment of the tunnel's over- or under-excavation situation, including: After the design model and the measured point cloud are matched, the spatial difference between each measured point cloud and the design model is calculated. This spatial difference is defined as the vertical distance from the measured point to the design model. A voxel model is constructed from the design model and the measured point cloud. The over-excavation and under-excavation volume of each triangular unit or voxel unit is calculated and summed to obtain the total over-excavation and under-excavation volume of the entire tunnel cross section. Output the over-excavation and under-excavation calculation results in a structured data format.

7. A rapid analysis system for tunnel over-excavation and under-excavation based on model and point cloud matching, characterized in that, include: The point cloud acquisition module is used to acquire tunnel point cloud data and preprocess it to obtain high-quality measured point cloud data. The feature point extraction module is used to construct the design model, extract points on the contour lines of each section in the design model to obtain the design point cloud; extract feature points from the design point cloud and the measured point cloud, and establish spatial index structures for the feature points respectively. The registration module is used to perform initial registration of feature points between the design point cloud and the measured point cloud based on the spatial index results. It also uses an improved ICP algorithm to optimize the matching accuracy, iteratively finds the nearest corresponding point between the measured point cloud and the design model, and minimizes the distance error between them. During the matching process, the matching error is monitored in real time, and error correction and dynamic optimization are performed. Finally, the matching result is output. The over- and under-excavation volume quantification module is used to perform spatial difference analysis between the aligned measured point cloud and the design model based on the matching results, accurately calculate the over- and under-excavation volume of each point or region, and thus realize the quantitative assessment of the over- and under-excavation situation of the tunnel.

8. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the rapid analysis method for tunnel over-excavation and under-excavation based on model and point cloud matching as described in any one of claims 1-6.

9. A non-transitory computer-readable storage medium, characterized in that, The non-transitory computer-readable storage medium is used to store computer instructions, which, when executed by a processor, implement the rapid analysis method for tunnel over-excavation and under-excavation based on model and point cloud matching as described in any one of claims 1-6.

10. An electronic device, characterized in that, include: The device includes a processor, a memory, and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory to enable the electronic device to perform the rapid analysis method for tunnel over-excavation and under-excavation based on model and point cloud matching as described in any one of claims 1-6.

Images