Urban rail transit engineering structure health ultrasonic vision measurement method and device

Abstract

The invention discloses an urban rail transit engineering structure health ultrasonic vision measurement method and device, and relates to the technical field of urban rail transit engineering structure health monitoring. Firstly, the internal defect type and the internal defect position of a target structure are analyzed based on ultrasonic measurement data. By combining ultrasonic measurement and visual measurement, the internal defect and the external defect of the structure can be detected at the same time, and more comprehensive health state evaluation is provided. By analyzing the correlation between the internal defects and the external defects, the relation between the defects can be deeply understood, the root causes of the defects can be determined, the detected defects can be visually displayed in a three-dimensional image form by utilizing a three-dimensional reconstruction technology, and the detection accuracy is improved. And particularly, the associated defect three-dimensional diagram is subjected to three-dimensional display, so that engineers and decision makers can better understand the health condition of the structure, the workload of manual analysis is reduced, and the detection efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of urban rail transit engineering structure health monitoring, and in particular to an urban rail transit engineering structure health ultrasonic visual measurement method and device. Background Art

[0002] With the rapid development of urban rail transit, the demand for health monitoring of track structures is increasing.

[0003] Traditional monitoring methods mainly include manual inspection, mechanical monitoring and optical monitoring. Manual inspection involves regular visual inspection of the track structure by professional technicians and simple tool-assisted inspection, which has the problems of low efficiency, great influence by subjective factors and inability to monitor in real time.

[0004] Mechanical monitoring method refers to the use of mechanical sensors (such as strain gauges, displacement sensors, etc.) to monitor the deformation and stress of the track. The disadvantages are high installation and maintenance costs, sensors are easily affected by the environment and wear, and there are monitoring blind spots;

[0005] Optical monitoring refers to the use of laser scanning, digital photogrammetry and other technologies to monitor the geometric state of the track. The data processing is complex and is greatly affected by light and climate conditions.

[0006] Therefore, developing a new monitoring method that can monitor the health status of the track structure in real time and accurately is of great significance to ensure the safe operation of rail transit. Summary of the invention

[0007] In order to solve the above technical problems of track structure health monitoring, the present invention provides a method and device for ultrasonic visual measurement of urban rail transit engineering structure health. The following technical solutions are adopted:

[0008] An ultrasonic visual measurement method for the health of an urban rail transit engineering structure comprises the following steps:

[0009] Step 1, using an electric trolley to move along the side of the target track project;

[0010] Step 2: The ultrasonic measurement module and the visual measurement module installed on the electric vehicle simultaneously perform ultrasonic measurement and visual measurement on the target surface;

[0011] Step 3: The data processing module analyzes the ultrasonic measurement data and the visual measurement data respectively, firstly analyzing the internal defect type and internal defect position of the target structure based on the ultrasonic measurement data, and then analyzing the external defect type and external defect position of the target surface based on the visual measurement data;

[0012] Step 4, the data processing module performs time synchronization and spatial registration of the internal defect position and the external defect position;

[0013] Step 5: Analyze the correlation of internal defects and external defects that meet the spatial registration after time synchronization, set a correlation threshold, and determine internal defects and external defects with correlations greater than the correlation threshold as associated defects;

[0014] Step 6: Reconstruct the three-dimensional image of the associated defect using the three-dimensional reconstruction technology, and reconstruct other unassociated internal defects and external defects in three dimensions using the three-dimensional image of the associated defect as the visual center to generate a three-dimensional measurement image of the target structure.

[0015] By adopting the above technical solution, using an electric trolley to move along the side of the track and combining time synchronization and space registration technology, the position of the defect can be accurately located to improve the accuracy of detection.

[0016] By combining ultrasonic and visual measurements, both internal and external defects of the structure can be detected simultaneously, providing a more comprehensive health status assessment.

[0017] By analyzing the correlation between internal and external defects, we can have a deeper understanding of the connection between defects, which helps to determine the root cause of the defects. Using 3D reconstruction technology, the detected defects can be intuitively displayed in the form of 3D images. In particular, the 3D display of the associated defect 3D graph helps engineers and decision makers better understand the health status of the structure. The data processing module can automatically parse the measurement data, reducing the workload of manual analysis and improving detection efficiency.

[0018] Timely and accurate detection and location of internal and external defects in track structures helps prevent potential safety accidents and ensure the safety of passengers and staff. By detecting defects early, they can be repaired before they expand, thus reducing maintenance costs and potential larger-scale repair expenses.

[0019] Optionally, in step 4, the method for performing time synchronization and spatial registration is:

[0020] Assume that the timestamp of the ultrasonic measurement data is T u , the timestamp of the visual measurement data is T V , time synchronization is achieved through the following formula:

[0021] T sync =T u +ΔT;

[0022] ΔT=T V -T u ;

[0023] Where T sync This is the time after time synchronization.

[0024] By adopting the above technical solution, the timestamp T of the ultrasonic measurement data is adjusted. u It is consistent with the timestamp T of the visual measurement data. V Alignment, so that the two measurement data can correspond in time for subsequent data processing and analysis.

[0025] Optionally, the internal defect position obtained by ultrasonic measurement and the external defect position obtained by visual measurement are respectively converted into coordinates in a global coordinate system, a registration algorithm is used to find an optimal transformation matrix between the ultrasonic measurement data and the visual measurement data, and the found transformation matrix is ​​applied to the ultrasonic measurement data so that the position of the internal defect and the position of the external defect are in the same coordinate system;

[0026] A distance threshold Δd is set. If the distance between the converted internal defect and the external defect is less than the distance threshold Δd, it is determined that the internal defect and the external defect are spatially aligned.

[0027] By adopting the above technical solution, spatial registration involves matching the internal defect position obtained by ultrasonic measurement with the external defect position obtained by visual measurement. Use a registration algorithm (such as iterative closest point algorithm ICP, feature matching, etc.) to find the best transformation matrix between the ultrasonic measurement data and the visual measurement data. Apply the found transformation matrix to the ultrasonic measurement data so that the position of the internal defect is in the same coordinate system as the position of the external defect.

[0028] Optionally, in step 5, the positions of the internal defect i and the external defect e are P i and P e , the correlation is calculated using the following formula:

[0029]

[0030] Where C(i, e) is the correlation, P ik represents the position of internal defect i in the kth dimension, P ek is the position of the external defect e in the kth dimension, is the position mean of the internal defect i in all dimensions, is the position mean of the external defect e in all dimensions, and n is the number of dimensions of the defect position.

[0031] By adopting the above technical solution, C(i, e) quantifies the strength and direction of the linear relationship between the internal defect i and the external defect e, P ik The dimensions can be spatial coordinates (e.g., x, y, z) or other quantifiable attributes. is calculated by adding the position values ​​across all dimensions and dividing by the number of dimensions n. The calculation method of same.

[0032] Optionally, set a correlation threshold θ r , if C(i, e) is greater than θ r , then the internal defects and external defects are judged as associated defects.

[0033] By adopting the above technical solution, it is possible to achieve spatial registration of internal defects and external defects, and analyze the correlation between them, so as to determine whether they are related defects.

[0034] Optionally, in step 6, a feature matching algorithm is used to find corresponding feature points between the ultrasonic measurement data and the visual measurement data, the intrinsic parameters of the camera are obtained through calibration, the ultrasonic data and the visual data are stereo corrected to ensure that they are in the same plane and aligned, the depth of each feature point is calculated using the triangulation principle, and a three-dimensional point cloud is generated based on the position of the feature point and the calculated depth. The point cloud data is used to associate the defect three-dimensional image through Poisson reconstruction.

[0035] Optionally, a three-dimensional coordinate system is constructed with the geometric center of the associated defect three-dimensional image as the origin, and the positional relationship between other unassociated internal defects and external defects and the geometric center is calculated. The three-dimensional features of other unassociated internal defects and external defects are merged into the three-dimensional reconstruction of the associated defect three-dimensional image based on the positional relationship to generate a three-dimensional image of the target structure measurement.

[0036] By adopting the above technical solutions, the use of feature matching algorithms improves the accuracy of feature point recognition. Combined with the principle of triangulation, the depth information of each feature point can be accurately calculated, providing key data for three-dimensional reconstruction. The generated three-dimensional point cloud can intuitively display the three-dimensional morphology of the structure. The Poisson reconstruction technology further optimizes the point cloud data and generates high-fidelity three-dimensional images, which helps to analyze structural defects more deeply. By reconstructing the three-dimensional images of associated defects, the shape, size and location of the defects can be more intuitively understood, providing important information for repair and reinforcement. The three-dimensional features of unassociated internal defects and external defects are merged into the three-dimensional reconstruction of the associated defects to generate a complete three-dimensional map of the target structure measurement, providing a comprehensive three-dimensional perspective for structural health assessment. The three-dimensional coordinate system constructed with the geometric center of the three-dimensional image of the associated defect as the origin helps to accurately locate and navigate during maintenance and inspection.

[0037] A device for ultrasonic visual measurement of the structural health of an urban rail transit project is used to implement a method for ultrasonic visual measurement of the structural health of an urban rail transit project. The measuring device comprises an electric trolley, an ultrasonic measurement module, a visual measurement module and a data processing module. The ultrasonic measurement module and the visual measurement module are respectively installed on the electric trolley, and the electric trolley is located at the side of a target track project and moves. The ultrasonic measurement module is used to perform ultrasonic measurement on the target track project, and the visual measurement module is used to perform visual measurement on the target track project. The data processing module is respectively communicated with the ultrasonic measurement module and the visual measurement module. The analysis program designed by the method of steps 3 to 6 of the method for ultrasonic visual measurement of the structural health of an urban rail transit project described in claim 7 analyzes the ultrasonic measurement data and the visual measurement data, and outputs a three-dimensional measurement graph of the target structure.

[0038] Optionally, the ultrasonic measurement module is an ultrasonic sensor array, and the visual measurement module is a visual camera.

[0039] Optionally, the data processing module includes a memory and a computer, the memory is wirelessly connected to the ultrasonic measurement module and the visual measurement module via a wireless communication module, the computer is communicatively connected to the memory, the computer is pre-installed with an analysis program designed using steps 3 to 6 of a method for ultrasonic visual measurement of the structural health of an urban rail transit project, and the analysis program is run to output a three-dimensional measurement image of the target structure.

[0040] In summary, the present invention includes at least one of the following beneficial technical effects:

[0041] The present invention can provide a method and device for ultrasonic visual measurement of the health of urban rail transit engineering structures. By combining ultrasonic measurement and visual measurement, internal and external defects of the structure can be detected simultaneously, providing a more comprehensive health status assessment.

[0042] By analyzing the correlation between internal and external defects, we can have a deeper understanding of the connection between defects, which helps to determine the root cause of the defects. Using 3D reconstruction technology, the detected defects can be intuitively displayed in the form of 3D images. In particular, the 3D display of the associated defect 3D graph helps engineers and decision makers better understand the health status of the structure. The data processing module can automatically parse the measurement data, reducing the workload of manual analysis and improving detection efficiency.

[0043] Timely and accurate detection and location of internal and external defects in track structures helps prevent potential safety accidents and ensure the safety of passengers and staff. By detecting defects early, they can be repaired before they expand, thus reducing maintenance costs and potential larger-scale repair expenses. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 The present invention is a flow chart of a method for ultrasonic visual measurement of the health of an urban rail transit engineering structure. DETAILED DESCRIPTION

[0045] The present invention is further described in detail below in conjunction with the accompanying drawings.

[0046] The embodiment of the invention discloses a method and device for ultrasonic visual measurement of the structural health of an urban rail transit project.

[0047] Reference Figure 1 , Embodiment 1, a method for ultrasonic visual measurement of the health of an urban rail transit engineering structure, comprising the following steps:

[0048] Step 1, using an electric trolley to move along the side of the target track project;

[0049] Step 2: The ultrasonic measurement module and the visual measurement module installed on the electric vehicle simultaneously perform ultrasonic measurement and visual measurement on the target surface;

[0050] Step 3: The data processing module analyzes the ultrasonic measurement data and the visual measurement data respectively, firstly analyzing the internal defect type and internal defect position of the target structure based on the ultrasonic measurement data, and then analyzing the external defect type and external defect position of the target surface based on the visual measurement data;

[0051] Step 4, the data processing module performs time synchronization and spatial registration of the internal defect position and the external defect position;

[0052] Step 5: Analyze the correlation of internal defects and external defects that meet the spatial registration after time synchronization, set a correlation threshold, and determine internal defects and external defects with correlations greater than the correlation threshold as associated defects;

[0053] Step 6: Reconstruct the three-dimensional image of the associated defect using the three-dimensional reconstruction technology, and reconstruct other unassociated internal defects and external defects in three dimensions using the three-dimensional image of the associated defect as the visual center to generate a three-dimensional measurement image of the target structure.

[0054] By using an electric trolley to move along the side of the track and combining time synchronization and spatial registration technology, the position of defects can be accurately located, improving the accuracy of detection.

[0055] By combining ultrasonic and visual measurements, both internal and external defects of the structure can be detected simultaneously, providing a more comprehensive health status assessment.

[0056] By analyzing the correlation between internal and external defects, we can have a deeper understanding of the connection between defects, which helps to determine the root cause of the defects. Using 3D reconstruction technology, the detected defects can be intuitively displayed in the form of 3D images. In particular, the 3D display of the associated defect 3D graph helps engineers and decision makers better understand the health status of the structure. The data processing module can automatically parse the measurement data, reducing the workload of manual analysis and improving detection efficiency.

[0057] Timely and accurate detection and location of internal and external defects in track structures helps prevent potential safety accidents and ensure the safety of passengers and staff. By detecting defects early, they can be repaired before they expand, thus reducing maintenance costs and potential larger-scale repair expenses.

[0058] In step 4 of embodiment 2, the method for performing time synchronization and spatial registration is:

[0059] Assume that the timestamp of the ultrasonic measurement data is T u , the timestamp of the visual measurement data is T V , time synchronization is achieved through the following formula:

[0060] T sync =T u +ΔT;

[0061] ΔT=T V -T u ;

[0062] Where T sync This is the time after time synchronization.

[0063] Adjust the timestamp T of the ultrasonic measurement data u It is consistent with the timestamp T of the visual measurement data. V Alignment, so that the two measurement data can correspond in time for subsequent data processing and analysis.

[0064] The internal defect positions obtained by ultrasonic measurement and the external defect positions obtained by visual measurement are respectively converted into coordinates in the global coordinate system, and the registration algorithm is used to find the best transformation matrix between the ultrasonic measurement data and the visual measurement data. The found transformation matrix is ​​applied to the ultrasonic measurement data so that the positions of the internal defects and the external defects are in the same coordinate system;

[0065] In embodiment 3, a distance threshold Δd is set. If the distance between the converted internal defect and the external defect is less than the distance threshold Δd, it is determined that the internal defect and the external defect are spatially aligned.

[0066] Spatial registration involves matching the internal defect location obtained by ultrasonic measurement with the external defect location obtained by visual measurement. Use a registration algorithm (such as iterative closest point algorithm ICP, feature matching, etc.) to find the best transformation matrix between the ultrasonic measurement data and the visual measurement data. Apply the found transformation matrix to the ultrasonic measurement data so that the location of the internal defect is in the same coordinate system as the location of the external defect.

[0067] In Example 4, in step 5, the positions of the internal defect i and the external defect e are respectively i and P e , the correlation is calculated using the following formula:

[0068]

[0069] Where C(i, e) is the correlation, P ik represents the position of internal defect i in the kth dimension, P ek is the position of the external defect e in the kth dimension, is the position mean of the internal defect i in all dimensions, is the position mean of the external defect e in all dimensions, and n is the number of dimensions of the defect position.

[0070] C(i, e) quantifies the strength and direction of the linear relationship between internal defect i and external defect e, P ik The dimensions can be spatial coordinates (e.g., x, y, z) or other quantifiable attributes. is calculated by adding the position values ​​across all dimensions and dividing by the number of dimensions n. The calculation method of same.

[0071] Example 5, setting the correlation threshold θ r , if C(i, e) is greater than θ r , then the internal defects and external defects are judged as associated defects.

[0072] It is possible to realize spatial registration of internal defects and external defects, and analyze the correlation between them to determine whether they are related defects.

[0073] In Example 6, in step 6, a feature matching algorithm is used to find corresponding feature points between ultrasonic measurement data and visual measurement data, the intrinsic parameters of the camera are obtained through calibration, the ultrasonic data and the visual data are stereo corrected to ensure that they are in the same plane and aligned, the depth of each feature point is calculated using the principle of triangulation, and a three-dimensional point cloud is generated based on the position of the feature point and the calculated depth. The point cloud data is used to associate the defect three-dimensional image through Poisson reconstruction.

[0074] Example 7: Construct a three-dimensional coordinate system with the geometric center of the associated defect three-dimensional image as the origin, calculate the positional relationship between other unassociated internal defects and external defects and the geometric center, and merge the three-dimensional features of other unassociated internal defects and external defects into the three-dimensional reconstruction of the associated defect three-dimensional image based on the positional relationship to generate a three-dimensional image of the target structure measurement.

[0075] The use of feature matching algorithms improves the accuracy of feature point recognition. Combined with the principle of triangulation, the depth information of each feature point can be accurately calculated, providing key data for three-dimensional reconstruction. The generated three-dimensional point cloud can intuitively display the three-dimensional morphology of the structure. Poisson reconstruction technology further optimizes the point cloud data and generates high-fidelity three-dimensional images, which helps to analyze structural defects more deeply. By reconstructing the three-dimensional image of the associated defects, the shape, size and location of the defects can be more intuitively understood, providing important information for repair and reinforcement. The three-dimensional features of unassociated internal defects and external defects are merged into the three-dimensional reconstruction of the associated defects to generate a complete three-dimensional map of the target structure measurement, providing a comprehensive three-dimensional perspective for structural health assessment. The three-dimensional coordinate system constructed with the geometric center of the three-dimensional image of the associated defects as the origin facilitates accurate positioning and navigation during maintenance and inspection.

[0076] Embodiment 8, an ultrasonic visual measurement device for the structural health of an urban rail transit project, used to implement an ultrasonic visual measurement method for the structural health of an urban rail transit project, the measurement device includes an electric trolley, an ultrasonic measurement module, a visual measurement module and a data processing module, the ultrasonic measurement module and the visual measurement module are respectively installed on the electric trolley, the electric trolley is located at the side of the target track project and moves, the ultrasonic measurement module is used to perform ultrasonic measurement on the target track project, the visual measurement module is used to perform visual measurement on the target track project, the data processing module is communicated with the ultrasonic measurement module and the visual measurement module respectively, and an analysis program designed by the method of steps 3 to step 6 of the ultrasonic visual measurement method for the structural health of an urban rail transit project of claim 7 is used to analyze the ultrasonic measurement data and the visual measurement data, and output a three-dimensional measurement graph of the target structure.

[0077] In specific implementation, the electric cart can be an AGV electric cart, the battery of the AGV electric cart can power the ultrasonic measurement module and the visual measurement module, the ultrasonic measurement module and the visual measurement module can be installed through an angle adjustment mechanism such as an electric pan-tilt head, so that the ultrasonic measurement module and the visual measurement module can be aligned with the same position.

[0078] In Example 9, the ultrasonic measurement module is an ultrasonic sensor array, and the visual measurement module is a visual camera.

[0079] Embodiment 10, the data processing module includes a memory and a computer, the memory is wirelessly connected to the ultrasonic measurement module and the visual measurement module through the wireless communication module, the computer is communicatively connected to the memory, the computer is pre-installed with an analysis program designed using steps 3 to 6 of a method for ultrasonic visual measurement of the health of urban rail transit engineering structures, and the analysis program is run to output a three-dimensional measurement image of the target structure.

[0080] The following is a specific example to illustrate the implementation principle of a method and device for ultrasonic visual measurement of the health of urban rail transit engineering structures of the present invention:

[0081] A subway tunnel in a certain city has been in operation for many years. To ensure the safety of the tunnel structure, it is necessary to conduct regular structural health inspections on the tunnel. The method and device provided by the present invention are used to perform ultrasonic visual measurement on the subway tunnel to evaluate the tunnel health.

[0082] An AGV electric trolley is used to move along the side of the subway tunnel. The AGV electric trolley is battery-powered to ensure a smooth measurement process.

[0083] An ultrasonic sensor array and a visual camera are installed on the electric trolley to perform ultrasonic and visual measurements of the tunnel wall simultaneously.

[0084] The data processing module analyzes the ultrasonic measurement data and the visual measurement data respectively. First, the internal defect type and location of the tunnel are analyzed based on the ultrasonic measurement data, and then the external defect type and location of the tunnel are analyzed based on the visual measurement data.

[0085] Perform time synchronization and spatial registration. Assume that the timestamp of the ultrasound measurement data is T1, and the timestamp of the visual measurement data is T2. Convert the ultrasound measurement data and the visual measurement data into coordinates in the global coordinate system, and use the ICP algorithm to find the optimal transformation matrix to achieve spatial registration.

[0086] Analyze the correlation between internal defects and external defects. Set a correlation threshold. If the calculated correlation is greater than the threshold, it is determined to be an associated defect.

[0087] The feature matching algorithm is used to find the corresponding feature points between the ultrasonic measurement data and the visual measurement data. The camera internal parameters are obtained through calibration, and stereo correction and triangulation are performed to generate a 3D point cloud. The Poisson reconstruction technology is used to generate a 3D image of the associated defect, and a 3D coordinate system is constructed with the geometric center of the 3D image of the associated defect as the origin. The 3D features of other unassociated internal and external defects are merged into the 3D image of the associated defect to generate a 3D map of the tunnel structure measurement.

[0088] The measuring device includes an electric trolley, an ultrasonic measuring module, a visual measuring module and a data processing module. The data processing module is connected with the ultrasonic measuring module and the visual measuring module through a wireless communication module, and runs an analysis program to output a three-dimensional measurement map of the tunnel structure.

[0089] Through this ultrasonic visual measurement, it was found that there were many internal and external defects in the subway tunnel. Among them, some internal defects were highly correlated with external defects and were determined to be associated defects. The generated three-dimensional tunnel structure measurement map intuitively displayed the location, shape and size of the defects, providing an important basis for subsequent maintenance and reinforcement. By discovering defects early, potential safety accidents were avoided and the safety of passengers and staff was guaranteed.

[0090] The above are all preferred embodiments of the present invention, and are not intended to limit the protection scope of the present invention. Therefore, any equivalent changes made based on the structure, shape, and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for ultrasonic visual measurement of the structural health of urban rail transit engineering, characterized in that: The following steps are involved: Step 1, using an electric trolley to move along the side of the target track project; Step 2: The ultrasonic measurement module and the visual measurement module installed on the electric vehicle simultaneously perform ultrasonic measurement and visual measurement on the target surface; Step 3: The data processing module analyzes the ultrasonic measurement data and the visual measurement data respectively, firstly analyzing the internal defect type and internal defect position of the target structure based on the ultrasonic measurement data, and then analyzing the external defect type and external defect position of the target surface based on the visual measurement data; Step 4, the data processing module performs time synchronization and spatial registration of the internal defect position and the external defect position; Step 5: Analyze the correlation of internal defects and external defects that meet the spatial registration after time synchronization, set a correlation threshold, and determine internal defects and external defects with correlations greater than the correlation threshold as associated defects; Step 6: Reconstruct the three-dimensional image of the associated defect using the three-dimensional reconstruction technology, and reconstruct other unassociated internal defects and external defects in three dimensions using the three-dimensional image of the associated defect as the visual center to generate a three-dimensional measurement image of the target structure.

2. The method for ultrasonic visual measurement of the structural health of an urban rail transit project according to claim 1, characterized in that: In step 4, the method for performing time synchronization and spatial registration is: Assume that the timestamp of the ultrasonic measurement data is T u , the timestamp of the visual measurement data is T V , time synchronization is achieved through the following formula: T sync =T u +ΔT; ΔT=T V -T u ; Where T sync This is the time after time synchronization.

3. The method for ultrasonic visual measurement of the structural health of an urban rail transit project according to claim 2 is characterized in that: The internal defect positions obtained by ultrasonic measurement and the external defect positions obtained by visual measurement are respectively converted into coordinates in the global coordinate system, and the registration algorithm is used to find the best transformation matrix between the ultrasonic measurement data and the visual measurement data. The found transformation matrix is ​​applied to the ultrasonic measurement data so that the positions of the internal defects and the external defects are in the same coordinate system; A distance threshold Δd is set. If the distance between the converted internal defect and the external defect is less than the distance threshold Δd, it is determined that the internal defect and the external defect are spatially aligned.

4. The method for ultrasonic visual measurement of the structural health of an urban rail transit project according to claim 3 is characterized in that: In step 5, let the positions of internal defect i and external defect e be P i and P e , the correlation is calculated using the following formula: Where C(i, e) is the correlation, P ik represents the position of internal defect i in the kth dimension, P ek is the position of the external defect e in the kth dimension, is the position mean of the internal defect i in all dimensions, is the position mean of the external defect e in all dimensions, and n is the number of dimensions of the defect position.

5. The method for ultrasonic visual measurement of the structural health of an urban rail transit project according to claim 4, characterized in that: Set the correlation threshold θ r , if C(i, e) is greater than θ r , then the internal defects and external defects are judged as associated defects.

6. The method for ultrasonic visual measurement of the structural health of an urban rail transit project according to claim 5, characterized in that: In step 6, a feature matching algorithm is used to find the corresponding feature points between the ultrasonic measurement data and the visual measurement data. The intrinsic parameters of the camera are obtained through calibration, and the ultrasonic data and the visual data are stereo corrected to ensure that they are in the same plane and aligned. The depth of each feature point is calculated using the triangulation principle. According to the position of the feature point and the calculated depth, a three-dimensional point cloud is generated. The point cloud data is used to associate the defect three-dimensional image through Poisson reconstruction.

7. The method for ultrasonic visual measurement of the structural health of an urban rail transit project according to claim 6, characterized in that: A three-dimensional coordinate system is constructed with the geometric center of the associated defect three-dimensional image as the origin, and the positional relationship between other unassociated internal defects and external defects and the geometric center is calculated. The three-dimensional features of other unassociated internal defects and external defects are merged into the three-dimensional reconstruction of the associated defect three-dimensional image based on the positional relationship to generate a three-dimensional image of the target structure measurement.

8. An ultrasonic visual measurement device for the structural health of urban rail transit engineering, characterized in that: Used to implement the ultrasonic visual measurement method for the structural health of an urban rail transit project described in claim 7, the measuring device includes an electric trolley, an ultrasonic measurement module, a visual measurement module and a data processing module, the ultrasonic measurement module and the visual measurement module are respectively installed on the electric trolley, the electric trolley is located at the side of the target track project and moves, the ultrasonic measurement module is used to perform ultrasonic measurement on the target track project, the visual measurement module is used to perform visual measurement on the target track project, the data processing module is communicated with the ultrasonic measurement module and the visual measurement module respectively, and an analysis program designed by the method of steps 3 to 6 of the ultrasonic visual measurement method for the structural health of an urban rail transit project described in claim 7 is used to analyze the ultrasonic measurement data and the visual measurement data, and output a three-dimensional measurement graph of the target structure.

9. The ultrasonic visual measurement device for urban rail transit engineering structure health according to claim 8, characterized in that: The ultrasonic measurement module is an ultrasonic sensor array, and the visual measurement module is a visual camera.

10. The ultrasonic visual measurement device for urban rail transit engineering structure health according to claim 9, characterized in that: The data processing module includes a memory and a computer. The memory is wirelessly connected to the ultrasonic measurement module and the visual measurement module via a wireless communication module. The computer is communicatively connected to the memory. The computer is pre-installed with an analysis program designed using steps 3 to 6 of the method for ultrasonic visual measurement of the health of urban rail transit engineering structures as described in claim 7. The analysis program is run to output a three-dimensional measurement image of the target structure.

Images