A method, device, equipment and medium for detecting damage of a concrete tubular structure

Abstract

The application discloses a damage detection method, device, equipment and medium of a concrete tubular structure, and relates to the technical field of structure health monitoring. The application first acquires multi-view images of the concrete tubular structure in a current state; performs three-dimensional reconstruction on the multi-view images to obtain a three-dimensional point cloud model; then performs spatial alignment and registration of the three-dimensional point cloud model and a standard reference model to obtain a three-dimensional point cloud model after spatial alignment and registration; compares and analyzes the three-dimensional point cloud model after spatial alignment and registration and the standard reference model to obtain a damage detection result of the concrete tubular structure. Based on the curved surface three-dimensional reconstruction technology, the application performs three-dimensional reconstruction on the multi-view images of the concrete tubular structure in the current state, can realize registration with the standard reference model through a global alignment strategy, replaces a traditional registration algorithm sensitive to an initial position, and realizes high-precision, high-efficiency and low-cost detection of the damage of the concrete tubular structure.

Description

Technical Field

[0001] This application relates to the field of structural health monitoring technology, and in particular to a method, apparatus, equipment and medium for damage detection of concrete tubular structures. Background Technology

[0002] Concrete tubular structures serve as critical infrastructure for urban water supply and water conservancy projects. However, due to long-term exposure to high internal pressure and the corrosive effects of a complex external environment, these pipes are highly susceptible to cumulative damage such as steel wire breakage and concrete deterioration. These potential defects can escalate into catastrophic pipe bursts. Pipe bursts not only cause widespread water outages, disrupting production and daily life, but can also trigger secondary disasters such as road collapses, resulting in enormous economic losses and social impacts.

[0003] Currently, traditional detection methods such as laser scanning and ultrasonic testing, while offering high accuracy, suffer from expensive equipment, complex procedures, and often require on-site operation by specialized technicians, hindering large-scale promotion and widespread adoption. Furthermore, traditional image-based 3D reconstruction algorithms often struggle to accurately match feature points when dealing with curved surfaces like concrete pipes, leading to reconstruction failures or severely inaccurate models.

[0004] Even if point cloud data is successfully acquired, the subsequent alignment and registration process still faces bottlenecks. Traditional alignment algorithms, such as the iterative nearest point method, are highly sensitive to the initial position and pose of the point cloud. Without good initial values, they are prone to getting trapped in local optima, leading to incorrect alignment results and severely impacting the accuracy of the final damage assessment. Summary of the Invention

[0005] The purpose of this application is to provide a method, apparatus, equipment and medium for detecting damage to concrete tubular structures, so as to improve detection accuracy and efficiency, and achieve high-precision, high-efficiency and low-cost detection of damage to concrete tubular structures.

[0006] To achieve the above objectives, this application provides the following solution.

[0007] In a first aspect, this application provides a method for damage detection of concrete tubular structures, including:

[0008] Acquire multi-view images of a concrete tubular structure in its current state;

[0009] Three-dimensional reconstruction is performed on multi-view images of a concrete tubular structure in its current state to obtain a three-dimensional point cloud model.

[0010] The three-dimensional point cloud model is spatially aligned and registered with a standard reference model to obtain a spatially aligned and registered three-dimensional point cloud model; the standard reference model is obtained based on multi-view images of a concrete tubular structure in an undamaged state.

[0011] By comparing and analyzing the spatially aligned and registered 3D point cloud model with the standard reference model, the damage detection results of the concrete tubular structure are obtained.

[0012] Optionally, the standard reference model can be obtained in the following way:

[0013] Three-dimensional reconstruction is performed on multi-view images of a concrete tubular structure in an undamaged state to obtain a standard three-dimensional point cloud model.

[0014] The standard 3D point cloud model is processed to generate a standard reference model.

[0015] Optionally, the standard 3D point cloud model is processed to generate a standard reference model, specifically including:

[0016] The standard three-dimensional point cloud model is projected onto a two-dimensional plane along the axis of the standard three-dimensional point cloud model to obtain two-dimensional point cloud data;

[0017] The random sample consensus algorithm is used to fit a circle to the two-dimensional point cloud data to obtain the center coordinates and radius of the fitted circle;

[0018] A digital cylinder is constructed based on the center coordinates and radius of the fitted circle, serving as a standard reference model.

[0019] Optionally, the 3D point cloud model is spatially aligned and registered with a standard reference model to obtain a spatially aligned and registered 3D point cloud model, specifically including:

[0020] The least squares method is used to perform plane fitting on the 3D point cloud data in the 3D point cloud model to obtain the fitting plane;

[0021] Calculate the rotation angle and rotation axis between the normal vector of the fitted plane and the principal axis of the global coordinate system; the principal axis of the global coordinate system is consistent with the principal axis of the standard reference model;

[0022] Construct a rotation matrix based on the rotation angle and rotation axis;

[0023] Based on the rotation matrix, spatial alignment and registration are performed on the 3D point cloud data in the 3D point cloud model to obtain the spatially aligned and registered 3D point cloud model.

[0024] Optionally, the rotation matrix is:

[0025] ;

[0026] in, For rotation matrix, It is the identity matrix. For rotation angle, It is an antisymmetric matrix for the rotational axis.

[0027] Optionally, the 3D point cloud model is spatially aligned and registered with a standard reference model to obtain a spatially aligned and registered 3D point cloud model, specifically including:

[0028] The farthest point sampling method is used to select multiple sampling points on the edge of the 3D point cloud model and determine the corresponding point of each sampling point in the standard reference model;

[0029] Based on multiple sampling points and the corresponding points of each sampling point, the objective function is solved to obtain the optimal rotation angle and optimal translation amount; the objective function is:

[0030] ;

[0031] in, To achieve the optimal rotation angle, The optimal translation amounts in the three directions are... and These are the rotation matrix and the translation matrix, respectively. For the 3D point cloud model The coordinates of each sampling point For the first The coordinates of the corresponding points of each sampling point in the standard reference model;

[0032] Based on the optimal rotation angle and optimal translation amount, the optimal rotation matrix and optimal translation matrix are determined as follows: and ;in, and These are the optimal rotation matrix and the optimal translation matrix, respectively. It is the identity matrix. The antisymmetric matrix of the rotational axis, with superscript Indicates transpose;

[0033] Based on the optimal rotation matrix, spatial alignment and registration are performed on the 3D point cloud data in the 3D point cloud model to obtain the spatially aligned and registered 3D point cloud model.

[0034] Optionally, the spatially aligned and registered 3D point cloud model is compared and analyzed with a standard reference model to obtain damage detection results for the concrete tubular structure, specifically including:

[0035] The radial deformation of each point in the spatially aligned and registered 3D point cloud model is calculated using the following formula, and the deformation and damage area of ​​the concrete tubular structure is identified based on the radial deformation.

[0036] ;

[0037] in, In the spatially aligned and registered 3D point cloud model, the first Radial deformation at each point In the spatially aligned and registered 3D point cloud model, the first The radius at each point The radius of the standard reference model;

[0038] The local surface roughness of each point in the spatially aligned and registered 3D point cloud model is calculated using the following formula, and the delamination damage area of ​​the concrete tubular structure is identified based on the surface roughness.

[0039] ;

[0040] in, In the spatially aligned and registered 3D point cloud model, the first Local surface roughness at a point, In the spatially aligned and registered 3D point cloud model, the first The first point in the neighborhood of the nth point One point, In the spatially aligned and registered 3D point cloud model, the first The first point in the neighborhood of the nth point The point at the th The projection points of the local reference plane of each point. In the spatially aligned and registered 3D point cloud model, the first The number of points in the neighborhood of each point;

[0041] The surface curvature of each point in the spatially aligned and registered 3D point cloud model is calculated using the following formula, and the expansion or depression damage area of ​​the concrete tubular structure is identified based on the surface curvature.

[0042] ;

[0043] in, In the spatially aligned and registered 3D point cloud model, the first The surface curvature at each point, , and These are the first three points in the 3D point cloud model after spatial alignment and registration. The eigenvalues ​​of the covariance matrix constructed from a point and its neighborhood points.

[0044] Secondly, this application provides a damage detection device for concrete tubular structures, wherein the damage detection device for concrete tubular structures applies the aforementioned damage detection method for concrete tubular structures, and the damage detection device for concrete tubular structures includes:

[0045] Image acquisition module is used to acquire multi-view images of the concrete tubular structure in its current state;

[0046] The 3D reconstruction module is used to perform 3D reconstruction on multi-view images of concrete tubular structures in their current state to obtain a 3D point cloud model.

[0047] The spatial alignment and registration module is used to perform spatial alignment and registration between the 3D point cloud model and the standard reference model to obtain the spatially aligned and registered 3D point cloud model; the standard reference model is obtained based on multi-view images of the concrete tubular structure in an undamaged state.

[0048] The detection module is used to compare and analyze the spatially aligned and registered 3D point cloud model with the standard reference model to obtain the damage detection results of the concrete tubular structure.

[0049] Thirdly, this application provides a computer device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-described method for detecting damage to concrete tubular structures.

[0050] Fourthly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-described method for detecting damage to concrete tubular structures.

[0051] According to the specific embodiments provided in this application, this application has the following technical effects.

[0052] This application provides a method, apparatus, equipment, and medium for damage detection of concrete tubular structures. The method first acquires multi-view images of the concrete tubular structure in its current state; then, it performs 3D reconstruction on these images to obtain a 3D point cloud model; next, it spatially aligns and registers the 3D point cloud model with a standard reference model to obtain a spatially aligned and registered 3D point cloud model; finally, it compares and analyzes the spatially aligned and registered 3D point cloud model with the standard reference model to obtain the damage detection results for the concrete tubular structure. This application, based on curved surface 3D reconstruction technology and multi-view images of the concrete tubular structure in its current state, achieves 3D reconstruction with a standard reference model through a global alignment strategy, replacing the traditional registration algorithm sensitive to initial position. This method can be combined with UAV photogrammetry technology to improve detection efficiency, forming a low-cost, easy-to-operate, and widely applicable new damage detection method. It has significant practical implications for timely identification of pipeline safety hazards and prevention of major accidents. Attached Figure Description

[0053] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0054] Figure 1 This is a flowchart illustrating a damage detection method for a concrete tubular structure according to an embodiment of this application.

[0055] Figure 2 This is a schematic diagram of a spatial alignment method provided in an embodiment of this application.

[0056] Figure 3 This is a schematic diagram of a registration method provided in an embodiment of this application.

[0057] Figure 4 A comparison diagram of the reconstruction method provided in one embodiment of this application and other methods.

[0058] Figure 5 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0059] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0060] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0061] In one exemplary embodiment, a damage detection method for concrete tubular structures is provided, such as... Figure 1 As shown, it includes the following steps 101-104.

[0062] Step 101: Obtain multi-view images of the concrete tubular structure in its current state.

[0063] Step 102: Perform 3D reconstruction on the multi-view images of the concrete tubular structure in its current state to obtain a 3D point cloud model.

[0064] Step 103: Spatially align and register the three-dimensional point cloud model with the standard reference model to obtain the spatially aligned and registered three-dimensional point cloud model; the standard reference model is obtained based on multi-view images of the concrete tubular structure in an undamaged state.

[0065] Step 104: Compare and analyze the spatially aligned and registered 3D point cloud model with the standard reference model to obtain the damage detection results of the concrete tubular structure.

[0066] This application is based on surface 3D reconstruction technology, which performs 3D reconstruction based on multi-view images of concrete tubular structures in their current state. It can achieve registration with the standard reference model through a global alignment strategy, replacing the traditional registration algorithm that is sensitive to the initial position. This method can be combined with UAV photogrammetry technology to improve detection efficiency and form a new damage detection method that is low-cost, easy to operate, and widely applicable. It has important practical significance for timely investigation of pipeline safety hazards and prevention of major accidents.

[0067] In one specific implementation of steps 101-104 above, firstly, multi-view images of the concrete pipe surface are acquired. Then, high-precision 3D point cloud models before and after damage are reconstructed using photogrammetry principles. Next, an accurate standard reference model is established by performing cylindrical fitting based on the Random Sample Consensus (RANSAC) algorithm on the intact point cloud model before damage. Subsequently, an improved Iterative Closest Point (FPS-ICP) algorithm is used to accurately register the point cloud model after damage with the reference model. Finally, by comparing the two models, a comprehensive quantitative assessment and visualization of the structural damage of the concrete pipe is performed from three dimensions: radial deformation, damage area identification based on surface roughness, and damage mode analysis based on surface curvature.

[0068] In another exemplary embodiment, step 101 above is the data acquisition process. An operator uses a camera or a smartphone with a high-resolution camera to walk around the concrete pipe segment to be inspected, taking pictures of the pipe surface from different angles and distances. During shooting, it is necessary to ensure that there is at least 60% overlap between adjacent photos, and that image feature points (such as surface texture, marking lines, etc.) are evenly distributed and clear, to ensure the success rate and accuracy of subsequent 3D reconstruction. In this embodiment, image acquisition is performed before the pipeline pressure failure test (intact state) and after the failure (damaged state).

[0069] In another exemplary embodiment, step 102 above is the 3D reconstruction process. Since a standard reference model needs to be applied in subsequent registration and analysis, the two sets of images (intact and damaged) are wirelessly transmitted to the computer in this step. Using a self-developed program, the Structure from Motion (SFM) algorithm is first used to calculate the camera pose (position and orientation) of each photo based on the matching feature points between the images, and a sparse point cloud is generated. Subsequently, the Multi-View Stereo (MVS) algorithm is used to perform dense matching based on the known camera pose, and a high-density 3D point cloud model of the concrete pipe surface is reconstructed.

[0070] In another exemplary embodiment, to facilitate subsequent measurements, the reconstructed 3D point cloud model needs to be spatially aligned, such as... Figure 2 As shown. In step 103 above, the global plane of the 3D point cloud data in the 3D point cloud model is fitted using the least squares method to obtain the fitted plane, and then the normal vector of the fitted plane is obtained. The loss function is... ,in, For loss function, , , The three-dimensional point cloud model are respectively the first sampling points axis, shaft and Axis coordinates , , The parameters are used for fitting. The rotation relationship between the normal vector and the principal axis of the global coordinate system is calculated, a rotation matrix is ​​generated, and this matrix is ​​applied to all points so that the central axis of the 3D point cloud model is approximately parallel to the principal axis of the global coordinate system.

[0071] The rotation matrix is ​​as follows:

[0072] ;

[0073] in, For rotation matrix, It is the identity matrix. For rotation angle, It is an antisymmetric matrix for the rotational axis.

[0074] In another exemplary embodiment, the standard reference model in steps 103 and 104 above is generated in the following manner.

[0075] Due to limitations in shooting angle or occlusion, the point cloud model in its intact state may contain missing data. To obtain complete geometric information of the concrete pipe, the aligned point cloud in its intact state is projected along the Z-axis onto the XY plane. Then, the Random Sample Consensus (RANSAC) algorithm is used to fit the projected points. This algorithm, through iterative random sampling, can effectively resist the influence of noise and local data gaps, ultimately finding the optimal fitted circle and determining its center coordinates and radius. Using these parameters, an ideal, smooth digital cylinder is generated as a standard reference model for subsequent damage comparison.

[0076] In another exemplary embodiment, to compare the differences between the 3D point cloud model and the reference model, they need to be registered. This embodiment employs an Iterative Nearest Point Algorithm (FPS-ICP) based on farthest point sampling, such as... Figure 3 As shown, the algorithm first selects a small number (e.g., 100) of representative points at the edges (such as the top and bottom) of the damaged point cloud model by sampling from the farthest points. Then, it performs ICP registration between these sampled points and the standard reference model. The registration process minimizes the objective function. To solve for the optimal rotation angle and the optimal translation in three directions, where, To achieve the optimal rotation angle, The optimal translation amounts in the three directions are... and These are the rotation matrix and the translation matrix, respectively. For the 3D point cloud model The coordinates of each sampling point For the first The coordinates of the corresponding points of each sampling point in the standard reference model are then used to further determine the optimal rotation matrix and optimal translation matrix based on the optimal rotation angle and optimal translation amount. and ;in, and These are the optimal rotation matrix and the optimal translation matrix, respectively. It is the identity matrix. The antisymmetric matrix of the rotational axis, with superscript This indicates transpose. Compared to the traditional ICP algorithm, which requires calculation for all points, this method improves computation speed by tens of times while maintaining registration accuracy. The reconstruction results of this application were compared and verified with the results of the 3D Gaussian Splatting (3DGS) method, such as... Figure 4 As shown, this application significantly outperforms the 3DGS method in both reconstruction accuracy and fit.

[0077] Using the coefficient of determination (R²) 2The degree of difference is quantified by R and Normalized Root Mean Square Error (NRMSE). 2 Defined by the following formula:

[0078] ;

[0079] In the formula, For the reconstruction results, The results of the laser scan. This represents the average value of the laser scan.

[0080] NRMSE is defined by the following formula:

[0081] ;

[0082] In the formula, To maximize the reconstruction result, This represents the maximum value of the laser scanning result. This represents the number of sampling points.

[0083] from Figure 4 As can be clearly seen, compared with the benchmark laser scanning data, the method of this application highly overlaps with it, exhibiting smooth contours and high consistency. The 3D Gaussian Splatting (3DGS) method shows several significant deviations and irregular fluctuations compared to the benchmark data. Furthermore, the R-value of the method of this application... 2 The value reached 0.96, higher than the 0.87 of the 3DGS method, indicating that its reconstruction results are more correlated with the true contour. At the same time, the NRMSE of the method in this application is only 3.52%, significantly lower than the 8.91% of the 3DGS method, indicating that the reconstruction error of this application is smaller and the accuracy is higher.

[0084] In another exemplary embodiment, a multi-dimensional damage assessment was performed in step 104 above. After registration, quantitative analysis was conducted at the following three levels:

[0085] 1. Deformation Analysis: To visually demonstrate the deformation, the cylindrical point cloud data is first transformed using coordinate transformation. Unfold into a two-dimensional planar diagram, where, For rotation angle, r Radius. Damage deformation. w By comparing the actual radius of each point in the damage point cloud with the radius of the standard reference model... R The difference was calculated to yield: ,in, In the spatially aligned and registered 3D point cloud model, the first Radial deformation at each point In the spatially aligned and registered 3D point cloud model, the first The radius at each point The radius is the standard reference model. This method can accurately measure deformations such as local bulging of the pipe wall caused by the fracture of prestressed tendons, and locate the point of maximum deformation.

[0086] 2. Damage Area Calculation: After the mortar protective layer peels off the surface of the concrete pipe, the exposed internal concrete or steel wire will roughen the surface. This characteristic can be used to identify the damage area. For the unfolded point cloud map, the local surface roughness of each point is calculated. The calculation method is as follows: Within the neighborhood of each point, a local reference plane is fitted using the least squares method, and then the average distance from all points in the neighborhood to this plane is calculated. The formula is: ,in, In the spatially aligned and registered 3D point cloud model, the first Local surface roughness at a point, In the spatially aligned and registered 3D point cloud model, the first The first point in the neighborhood of the nth point One point, In the spatially aligned and registered 3D point cloud model, the first The first point in the neighborhood of the nth point The point at the th The projection points of the local reference plane of each point. In the spatially aligned and registered 3D point cloud model, the first The number of points in the neighborhood of each point is calculated. A grayscale image of the calculated local surface roughness is generated, and a suitable threshold is set to identify areas exceeding the threshold as damaged or spalled areas. Finally, computer vision image processing algorithms are used to calculate the area of ​​the damaged region.

[0087] 3. Damage Mode Analysis: Surface curvature can effectively reflect more subtle damage modes. For the unfolded point cloud, Principal Component Analysis (PCA) is used to calculate the surface curvature of each point. Specifically, for each point and its neighborhood points, a covariance matrix is ​​constructed, and its eigenvalues ​​are used to... The formula for calculating surface curvature is as follows: ,in, In the spatially aligned and registered 3D point cloud model, the first The surface curvature at each point, , and These are the first three points in the 3D point cloud model after spatial alignment and registration. The eigenvalues ​​of the covariance matrix constructed from a point and its neighborhood. Generally, the greater the curvature, the more severe the damage. The calculated curvature values ​​are normalized (from 0 to 100%) to generate a damage level cloud map, thus visually displaying the distribution and severity of the damage.

[0088] In summary, this application, through a complete process, enables low-cost, high-efficiency, and high-precision damage detection of concrete pipes using ordinary smartphones.

[0089] Implementing the above steps can achieve the following technical effects.

[0090] (1) Low cost and easy to popularize: The embodiments of this application only require the use of widely popular smartphones and other devices as data acquisition devices, which greatly reduces the hardware cost of detection and makes large-scale, high-frequency pipeline inspection possible.

[0091] (2) High efficiency, portability and easy operation: The embodiments of this application utilize the portability of smartphones to enable testing personnel to easily reach various complex sites to collect data. The process is simple and can be mastered without professional training.

[0092] (3) Non-invasive high-precision detection: The embodiments of this application do not require contact with the tube body. Through three-dimensional point cloud analysis, millimeter-level precision measurement of tube surface deformation and damage can be achieved, which can accurately quantify the degree of damage.

[0093] (4) Multi-dimensional comprehensive evaluation: The embodiments of this application can not only detect deformation, but also identify and quantify different damage types such as mortar spalling and local bulging through roughness and curvature analysis, providing more comprehensive data support for pipeline maintenance decisions.

[0094] Based on the same inventive concept, this application also provides a damage detection device for concrete tubular structures to implement the damage detection method for concrete tubular structures described above. The solution provided by this device is similar to the solution described in the above method; therefore, the specific limitations of one or more embodiments of the damage detection device for concrete tubular structures provided below can be found in the limitations of the damage detection method for concrete tubular structures described above, and will not be repeated here.

[0095] In one exemplary embodiment, a damage detection device for a concrete tubular structure is provided, comprising:

[0096] Image acquisition module is used to acquire multi-view images of the concrete tubular structure in its current state;

[0097] The 3D reconstruction module is used to perform 3D reconstruction on multi-view images of concrete tubular structures in their current state to obtain a 3D point cloud model.

[0098] The spatial alignment and registration module is used to perform spatial alignment and registration between the 3D point cloud model and the standard reference model to obtain the spatially aligned and registered 3D point cloud model; the standard reference model is obtained based on multi-view images of the concrete tubular structure in an undamaged state.

[0099] The detection module is used to compare and analyze the spatially aligned and registered 3D point cloud model with the standard reference model to obtain the damage detection results of the concrete tubular structure.

[0100] In one exemplary embodiment, a computer device is provided, which may be a server or a terminal, and its internal structure diagram may be as follows. Figure 5 As shown, this computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements a damage detection method for concrete tubular structures.

[0101] Those skilled in the art will understand that Figure 5 The structures shown are merely block diagrams of some structures related to the present application and do not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than shown in the figures, or combine certain components, or have different component arrangements. In an exemplary embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0102] In one exemplary embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0103] In one exemplary embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0104] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0105] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0106] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0107] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0108] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for damage detection of concrete tubular structures, characterized in that, include: Acquire multi-view images of a concrete tubular structure in its current state; Three-dimensional reconstruction is performed on multi-view images of a concrete tubular structure in its current state to obtain a three-dimensional point cloud model. The three-dimensional point cloud model is spatially aligned and registered with a standard reference model to obtain a spatially aligned and registered three-dimensional point cloud model; the standard reference model is obtained based on multi-view images of a concrete tubular structure in an undamaged state. The spatially aligned and registered 3D point cloud model was compared and analyzed with a standard reference model to obtain damage detection results for the concrete tubular structure, specifically including: The radial deformation of each point in the spatially aligned and registered 3D point cloud model is calculated using the following formula, and the deformation and damage area of ​​the concrete tubular structure is identified based on the radial deformation. ； in, In the spatially aligned and registered 3D point cloud model, the first Radial deformation at each point In the spatially aligned and registered 3D point cloud model, the first The radius at each point The radius of the standard reference model; The local surface roughness of each point in the spatially aligned and registered 3D point cloud model is calculated using the following formula, and the delamination damage area of ​​the concrete tubular structure is identified based on the surface roughness. ； in, In the spatially aligned and registered 3D point cloud model, the first Local surface roughness at a point, In the spatially aligned and registered 3D point cloud model, the first Within the neighborhood of point n, the first One point, In the spatially aligned and registered 3D point cloud model, the first Within the neighborhood of point n, the first The point at the th The projection points of the local reference plane of each point. In the spatially aligned and registered 3D point cloud model, the first The number of points in the neighborhood of each point; The surface curvature of each point in the spatially aligned and registered 3D point cloud model is calculated using the following formula, and the expansion or depression damage area of ​​the concrete tubular structure is identified based on the surface curvature. ； in, In the spatially aligned and registered 3D point cloud model, the first The surface curvature at each point, , and These are the first three points in the 3D point cloud model after spatial alignment and registration. The eigenvalues ​​of the covariance matrix constructed from a point and its neighborhood points.

2. The damage detection method for concrete tubular structures according to claim 1, characterized in that, The standard reference model is obtained as follows: Three-dimensional reconstruction is performed on multi-view images of a concrete tubular structure in an undamaged state to obtain a standard three-dimensional point cloud model. The standard 3D point cloud model is processed to generate a standard reference model.

3. The damage detection method for concrete tubular structures according to claim 2, characterized in that, The standard 3D point cloud model is processed to generate a standard reference model, specifically including: The standard three-dimensional point cloud model is projected onto a two-dimensional plane along the axis of the standard three-dimensional point cloud model to obtain two-dimensional point cloud data; The random sample consensus algorithm is used to fit a circle to the two-dimensional point cloud data to obtain the center coordinates and radius of the fitted circle; A digital cylinder is constructed based on the center coordinates and radius of the fitted circle, serving as a standard reference model.

4. The damage detection method for concrete tubular structures according to claim 1, characterized in that, The 3D point cloud model is spatially aligned and registered with a standard reference model to obtain a spatially aligned and registered 3D point cloud model, specifically including: The least squares method is used to perform plane fitting on the 3D point cloud data in the 3D point cloud model to obtain the fitting plane; Calculate the rotation angle and rotation axis between the normal vector of the fitted plane and the principal axis of the global coordinate system; the principal axis of the global coordinate system is consistent with the principal axis of the standard reference model; Construct a rotation matrix based on the rotation angle and rotation axis; Based on the rotation matrix, spatial alignment and registration are performed on the 3D point cloud data in the 3D point cloud model to obtain the spatially aligned and registered 3D point cloud model.

5. The damage detection method for concrete tubular structures according to claim 4, characterized in that, The rotation matrix is: ； in, Let be a rotation matrix. It is the identity matrix. For rotation angle, It is an antisymmetric matrix for the rotational axis.

6. The damage detection method for concrete tubular structures according to claim 1, characterized in that, The 3D point cloud model is spatially aligned and registered with a standard reference model to obtain a spatially aligned and registered 3D point cloud model, specifically including: The farthest point sampling method is used to select multiple sampling points on the edge of the 3D point cloud model and determine the corresponding point of each sampling point in the standard reference model; Based on multiple sampling points and the corresponding points of each sampling point, the objective function is solved to obtain the optimal rotation angle and optimal translation amount; the objective function is: ； in, To achieve the optimal rotation angle, The optimal translation amounts in the three directions are... and These are the rotation matrix and the translation matrix, respectively. For the 3D point cloud model The coordinates of each sampling point For the first The coordinates of the corresponding points of each sampling point in the standard reference model; Based on the optimal rotation angle and optimal translation amount, the optimal rotation matrix and optimal translation matrix are determined as follows: and ;in, and These are the optimal rotation matrix and the optimal translation matrix, respectively. It is the identity matrix. The antisymmetric matrix of the rotational axis, with superscript Indicates transpose; Based on the optimal rotation matrix, spatial alignment and registration are performed on the 3D point cloud data in the 3D point cloud model to obtain the spatially aligned and registered 3D point cloud model.

7. A damage detection device for concrete tubular structures, characterized in that, The damage detection device for the concrete tubular structure applies the damage detection method for the concrete tubular structure according to any one of claims 1-6, and the damage detection device for the concrete tubular structure comprises: Image acquisition module is used to acquire multi-view images of the concrete tubular structure in its current state; The 3D reconstruction module is used to perform 3D reconstruction on multi-view images of concrete tubular structures in their current state to obtain a 3D point cloud model. The spatial alignment and registration module is used to perform spatial alignment and registration between the three-dimensional point cloud model and the standard reference model to obtain the spatially aligned and registered three-dimensional point cloud model; the standard reference model is obtained based on multi-view images of the concrete tubular structure in an undamaged state. The detection module is used to compare and analyze the spatially aligned and registered 3D point cloud model with the standard reference model to obtain the damage detection results of the concrete tubular structure.

8. A computer device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the computer program to implement the damage detection method for a concrete tubular structure according to any one of claims 1-6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the damage detection method for concrete tubular structures as described in any one of claims 1-6.

Images