A tunnel deformation inspection system and method based on monocular pose estimation

Abstract

This invention discloses a tunnel deformation inspection system and method based on monocular pose estimation. The system includes: a mobile monitoring platform for providing movement of the inspection system along the tunnel direction; an image acquisition device mounted on the upper part of the mobile monitoring platform via a pan-tilt unit, comprising three cameras rigidly connected in a T-shape, wherein the first and second cameras are used to acquire images of checkerboard targets at measuring points A and B on both sides of each tunnel section, respectively, and the third camera is used to acquire images of checkerboard targets at measuring point C on the tunnel arch of each section; an odometer for acquiring mileage data after the mobile monitoring platform enters the tunnel and determining the acquisition position of the mobile monitoring platform; a microcomputer terminal for controlling the mobile inspection platform to enter the tunnel; receiving mileage data provided by the odometer and controlling the image acquisition device to acquire images based on the mileage data; and calculating the convergence deformation of each section and the settlement of the arch based on the acquired images.

Description

Technical Field

[0001] This invention belongs to the field of civil engineering monitoring, and in particular relates to a tunnel deformation inspection system and method based on monocular pose estimation. Background Technology

[0002] Current tunnel deformation monitoring methods can be broadly categorized into two types: traditional manual monitoring and automated monitoring. Traditional manual monitoring relies on staff using tools such as dial indicators and convergence meters to sequentially inspect each monitoring section along the tunnel line, or on setting up deformation observation stations within the tunnel and employing precision measuring equipment such as theodolites, levels, and distance measuring instruments for measurement and data processing to understand tunnel deformation. These methods are highly dependent on manual labor, inefficient, and cannot avoid human error. Among common automated tunnel deformation monitoring methods, static levels and Basel convergence systems suffer from insufficient deformation information acquisition, only monitoring either convergence or settlement, and failing to achieve three-dimensional displacement monitoring. While 3D laser scanning, fiber optic sensing systems, and measuring robots can acquire richer deformation information, they are expensive, and achieving full-line inspection of a subway line requires significant financial investment. Summary of the Invention

[0003] To address the problems existing in the prior art, the purpose of this application is to provide a tunnel deformation inspection system and method based on monocular pose estimation.

[0004] According to a first aspect of the embodiments of this application, a tunnel deformation inspection system based on monocular pose estimation is provided, comprising:

[0005] A mobile monitoring platform is used to monitor the movement of the inspection system along the tunnel direction;

[0006] The image acquisition device is mounted on the upper part of the mobile monitoring platform via a pan-tilt unit. It includes three cameras rigidly connected in a T-shape. The first and second cameras are used to acquire images of checkerboard targets at measuring points A and B on both sides of each section of the tunnel, respectively. The third camera is used to acquire images of checkerboard targets at measuring point C on the arch of each section of the tunnel. The checkerboard targets at measuring points A, B, and C are preset.

[0007] An odometer is used to acquire mileage data after the mobile monitoring platform enters the tunnel and to determine the data collection location of the mobile monitoring platform.

[0008] The microcomputer terminal is used to control the mobile detection platform to enter the tunnel; receive mileage data provided by the odometer and control the image acquisition device to acquire images based on the mileage data; and calculate the convergence deformation of each section and the settlement of the arch based on the acquired images.

[0009] Furthermore, the formation process of the checkerboard target includes:

[0010] ArUco markers are embedded within the white squares of a checkerboard pattern to form a checkerboard target that combines ArUco markers with the checkerboard pattern. Each checkerboard target has an ArUco marker that records the mileage information of the section where the checkerboard target is located.

[0011] According to a second aspect of the embodiments of this application, a tunnel deformation inspection method based on monocular pose estimation is provided, implemented through a microcomputer terminal in the system described in the first aspect, the method comprising:

[0012] Step S1: Control the mobile monitoring platform to move to the first section, and use the images acquired by the image acquisition device to calibrate the internal parameters of the first camera, the second camera, and the third camera, as well as the external parameters between the cameras;

[0013] Step S2: Control the image acquisition device to acquire target-containing images at measuring points A, B, and C, and calculate the tunnel convergence deformation and arch settlement at the first section using the target-containing images;

[0014] Step S3: Control the mobile monitoring platform to move to the second section, and calculate the tunnel convergence deformation and arch settlement at the second section using the process in step S2;

[0015] Step S4: Repeat step S3 until the tunnel convergence deformation and arch settlement at all cross sections have been calculated, and then control the mobile monitoring platform to return.

[0016] Further, in step S1, controlling the mobile monitoring platform to move to the first cross-section and calibrating the internal parameters of the first camera, the second camera, and the third camera using images acquired by the image acquisition device includes:

[0017] Step S11: Control the movement of the mobile monitoring platform according to the real-time mileage data transmitted by the odometer. When the mileage data transmitted by the odometer matches the mileage information of the first section to be measured, control the mobile monitoring platform to stop moving. At this time, the mobile monitoring platform reaches the first section.

[0018] Step S12: Control the first camera, the second camera and the third camera to acquire several images of the target at measurement points A, B and C corresponding to the cross section respectively;

[0019] Step S13: Based on the three sets of images transmitted by each camera, extract feature points and use Zhang Zhengyou's calibration method to solve for the intrinsic parameter matrices M of the first camera, the second camera, and the third camera respectively. in1 M in2 M in3 , which serve as internal parameters for each camera.

[0020] Furthermore, in step S1, calibrating the external parameters between the cameras includes:

[0021] Step S14: Control the first camera and the second camera to acquire calibration image a1 and calibration image b1 respectively from the checkerboard target of measuring point A and measuring point B corresponding to the first cross section;

[0022] Step S15: Based on the calibration images a1 and b1, extract feature points. Using the internal parameters of each camera, employ the PnP algorithm to solve for the pose transformation matrix between the target at measurement points A and B corresponding to the first cross-section and the first and second cameras, respectively, under this pose.

[0023] Step S16: After controlling the gimbal to change the pose of each camera, repeat steps S14 and S15.

[0024] Step S17: Repeat step S16 several times to obtain multiple sets of pose transformation matrices, thereby solving for the extrinsic parameter matrix H from the first camera to the second camera. 12 ;

[0025] Step S18: Using the methods in steps S14 to S17, solve for the extrinsic parameter matrix H from the first camera to the third camera. 13 ;

[0026] Step S19: Using the methods in steps S14 to S17, solve for the extrinsic parameter matrix H from the second camera to the third camera. 23 .

[0027] Furthermore, the extrinsic parameter matrix from the first camera to the second camera is solved, including:

[0028] Step S171: Denote the coordinate systems of measuring point A, measuring point B, and the first and second cameras as O. A O B The coordinates of any point P on the checkerboard target at measuring point A, along with O1 and O2, in the above coordinate system are denoted as P1, O2 ... A P B Based on the rigid body transformation law, for each pose state (P1, P2), the following can be obtained:

[0029]

[0030] In the formula H BA Let H be the pose transformation matrix from measurement point B to measurement point A. A1 Let H be the extrinsic parameter matrix from measurement point A to the first camera. 2B Given the extrinsic parameter matrix from the second camera to measurement point B, we can derive:

[0031] H BA H2B H 12 H A1 =I

[0032] Step S172: Using the obtained multiple sets of pose transformation matrices, the following set of equations is obtained:

[0033]

[0034]

[0035] ...

[0036]

[0037] In the formula, the superscript (j) represents the data after the j-th pose change, 0≤j≤n, and n is the number of pose changes. Let be the external parameter matrix from the first camera to measurement point A under the initial pose. Let be the external parameter matrix from the second camera to measurement point B under the initial pose; let and Let these be the homogeneous transformation matrices of the motion trajectories of the first camera and the second camera from their initial poses to the i-th pose, respectively, thus yielding:

[0038] H2 (i) H 12 =H 12 H1 (i)

[0039] Step S173: Solve the system of equations in step S172 using the hand-eye calibration method to obtain the external parameter matrix H from the first camera to the second camera. 12 .

[0040] Further, step S2 includes:

[0041] Step S21: Control the first camera, the second camera and the third camera to acquire images of the checkerboard target on the corresponding measurement points A, B and C of the cross section, respectively, to obtain target image a, target image b and target image c;

[0042] Step S22: Using the target image a and target image b, extract feature points, select the target world coordinate system where the measurement point A is located as the main coordinate system, and use the external parameters between the cameras to calculate the coordinates (x, y, y) of the measurement point B in the main coordinate system according to the following formula. b y b , z b ):

[0043]

[0044] In the formula M exAM is the external parameter matrix of the checkerboard target from the first camera to the measuring point A corresponding to the first cross-section. exB This is the external parameter matrix H of the checkerboard target from the second camera to the measuring point B corresponding to the first cross-section. 12 This is the external parameter matrix from the first camera to the second camera;

[0045] Step S23: Using the target image a and target image c, extract feature points, select the target world coordinate system at measurement point A as the main coordinate system, and use the external parameters between the cameras to calculate the coordinates (x, y, c) of measurement point C in the main coordinate system according to the following formula. c y c , z c ):

[0046]

[0047] Formula M exC This is the external parameter matrix of the checkerboard target from the third camera to the measuring point C corresponding to the first cross section at this time;

[0048] Step S24: Using the coordinates of measuring points B and C in the main coordinate system, calculate the length of side AB and the height h on the base AB of the survey line triangle ABC according to the following formula:

[0049]

[0050]

[0051]

[0052] p = (AB + AC + BC) / 2

[0053]

[0054] Step S25: Based on the length of side AB and the height h on the bottom side AB, calculate the tunnel convergence deformation Δd and tunnel arch settlement ΔH of the first cross-section:

[0055] Δd=AB-A′B′

[0056] ΔH=hh′

[0057] In the formula, A′B′ is the length between measuring points A and B obtained in the previous inspection, and h′ is the height on the base AB of triangle ABC obtained in the previous inspection.

[0058] According to a third aspect of the embodiments of this application, a tunnel deformation inspection device based on monocular pose estimation is provided, implemented through a microcomputer terminal in the system described in the first aspect, the method comprising:

[0059] The calibration module is used to control the mobile monitoring platform to move to the first cross section and to calibrate the internal parameters of the first camera, the second camera and the third camera, as well as the external parameters between the cameras, using the images acquired by the image acquisition device.

[0060] The first calculation module is used to control the image acquisition device to acquire target-containing images at measuring points A, B, and C, and to calculate the tunnel convergence deformation and arch settlement at the first cross section through the target-containing images.

[0061] The second calculation module is used to control the mobile monitoring platform to move to the second section and calculate the tunnel convergence deformation and arch settlement at the second section using the process of the first calculation module.

[0062] The return module is used to repeat the process of the second calculation module until the tunnel convergence deformation and arch settlement at all cross sections are calculated, and then control the mobile monitoring platform to return.

[0063] According to a fourth aspect of the embodiments of this application, an electronic device is provided, comprising:

[0064] One or more processors;

[0065] Memory, used to store one or more programs;

[0066] When the one or more programs are executed by the one or more processors, the one or more processors perform the method as described in the second aspect.

[0067] According to a fifth aspect of the present application, a computer-readable storage medium is provided that stores computer instructions thereon, which, when executed by a processor, implement the steps of the method as described in the second aspect.

[0068] The technical solutions provided by the embodiments of this application may include the following beneficial effects:

[0069] This application uses a novel tunnel deformation inspection system for monitoring. Compared with existing monitoring systems, the hardware equipment of this application has good adaptability to field conditions, is relatively simple to prepare, is convenient and flexible to deploy, has low cost, stable performance, and is not easily damaged.

[0070] This application employs a novel tunnel deformation inspection system for monitoring. Compared to existing monitoring methods, the monitoring method of this application improves data acquisition efficiency and measurement accuracy, reduces reliance on manual labor, and obtains more comprehensive information on cross-sectional deformation. At the same time, it can monitor the convergence of tunnel cross-sections and the settlement deformation of the arch, which has great advantages in the field of measurement.

[0071] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description

[0072] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0073] Figure 1 This is a schematic diagram of a tunnel deformation inspection system based on monocular pose estimation.

[0074] Figure 2 This is a schematic diagram of the bottom of the tunnel deformation inspection system.

[0075] Figure 3 This is a schematic diagram of a ChArUco flat panel, which combines ArUco markers with a checkerboard pattern for self-identification.

[0076] Figure 4 This is a schematic diagram of the inspection process.

[0077] Figure 5 This is a diagram illustrating the principle of deformation calculation.

[0078] Figure 6 This is a block diagram of a tunnel deformation inspection device based on monocular pose estimation, according to an exemplary embodiment.

[0079] Figure 7 This is a schematic diagram of an electronic device according to an exemplary embodiment.

[0080] In the diagram: 1. Mobile monitoring platform; 2. Image acquisition device; 21. First camera; 22. Second camera; 23. Third camera; 3. Pan-tilt unit; 4. Odometer; 5. Checkerboard target. Detailed Implementation

[0081] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application.

[0082] The terminology used in this application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. The singular forms “a,” “the,” and “the” used in this application and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0083] It should be understood that although the terms first, second, third, etc., may be used in this application to describe various information, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, without departing from the scope of this application, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Depending on the context, the word "if" as used herein may be interpreted as "when," "when," or "in response to determination."

[0084] Coordinate system explanation:

[0085] The origins of the coordinate systems of the first camera 21, the second camera 22, and the third camera 23 are the optical centers O1, O2, and O3 of each camera, respectively. The origin of the coordinate system of side point A is the lower right corner of the checkerboard target 5 at that location. The origin of the coordinate system of measurement point B is the lower left corner of the checkerboard target 5 at that location, and the origin of the coordinate system of measurement point C is the lower right corner of the checkerboard target 5 at that location. The directions of each coordinate axis are as follows: Figure 5 As shown.

[0086] Example 1

[0087] This application provides a tunnel deformation inspection system based on monocular pose estimation, such as... Figure 1 and Figure 2 As shown, the system may include: a mobile monitoring platform 1, used to provide the movement of the inspection system along the tunnel direction; an image acquisition device 2, mounted on the upper part of the mobile monitoring platform 1 via a pan-tilt unit 3, the high-precision pan-tilt unit 3 allowing the upper device to rotate a certain angle in three dimensions, the image acquisition device 2 including three cameras rigidly connected in a T-shape, the first camera 21 and the second camera 22 being used to acquire images of checkerboard targets 5 at measuring points A and B on both sides of each section of the tunnel, respectively, and the third camera 23 being used to acquire images of checkerboard targets 5 at measuring point C on the arch of each section of the tunnel, wherein the checkerboard targets 5 at measuring points A, B, and C are pre-set; an odometer 4, used to acquire mileage data after the mobile monitoring platform 1 enters the tunnel and determine the acquisition position of the mobile monitoring platform 1; a microcomputer terminal, used to control the mobile monitoring platform to enter the tunnel; receive the mileage data provided by the odometer 4 and control the image acquisition device 2 to acquire images according to the mileage data, and calculate the convergence deformation of each section and the settlement of the arch based on the acquired images.

[0088] Specifically, such as Figure 3 As shown, the formation process of the chessboard target 5 includes:

[0089] ArUco markers are embedded within the white grid of the checkerboard pattern to form a self-identifying array ChArUco plate that combines ArUco markers with the checkerboard pattern. The ArUco marker on each checkerboard target 5 records the mileage information of the section where the checkerboard target 5 is located.

[0090] As can be seen from the above embodiments, this system uses a self-identifying array ChArUco flat plate that combines ArUco markers with a checkerboard pattern. This is beneficial for the rapid detection and identification of feature corner points on the target, while the mileage information of the section where the ArUco markers are located can facilitate the accurate positioning of the mobile monitoring platform 1.

[0091] Example 2

[0092] This application provides a flowchart of a tunnel deformation inspection method based on monocular pose estimation, such as... Figure 4 As shown, this method is implemented through a microcomputer terminal in the above system, combining a non-common field-of-view camera and monocular pose estimation, and may include the following steps:

[0093] Step S1: Control the mobile monitoring platform 1 to move to the first section, and use the image acquired by the image acquisition device 2 to calibrate the internal parameters of the first camera 21, the second camera 22 and the third camera 23 as well as the external parameters between the cameras;

[0094] Step S2: Control the image acquisition device 2 to acquire target-containing images at measuring points A, B, and C, and calculate the tunnel convergence deformation and arch settlement at the first section using the target-containing images;

[0095] Step S3: Control the mobile monitoring platform 1 to move to the second section, and calculate the tunnel convergence deformation and arch settlement at the second section using the process in step S2;

[0096] Step S4: Repeat step S3 until the tunnel convergence deformation and arch settlement at all cross sections have been calculated, and then control the mobile monitoring platform 1 to return.

[0097] In the specific implementation of step S1, the mobile monitoring platform 1 is controlled to move to the first section, and the images acquired by the image acquisition device 2 are used to calibrate the internal parameters of the first camera 21, the second camera 22 and the third camera 23 as well as the external parameters between the cameras.

[0098] Specifically, the process of calibrating the internal parameters of the first camera 21, the second camera 22, and the third camera 23 may include:

[0099] Step S11: Control the movement of the mobile monitoring platform 1 according to the real-time mileage data transmitted by the odometer 4. When the mileage data transmitted by the odometer 4 matches the mileage information of the first section to be measured, control the mobile monitoring platform 1 to stop moving. At this time, the mobile monitoring platform 1 reaches the first section.

[0100] Specifically, by matching the mileage data transmitted by the odometer 4 with the mileage information of the first section to be measured, the working position of the mobile monitoring platform 1 at this section is ensured to coincide with the working position at this section during the last inspection, thereby reducing measurement errors.

[0101] Step S12: Control the first camera 21, the second camera 22 and the third camera 23 to acquire several images of the target on the corresponding measurement points A, B and C of the cross section.

[0102] Step S13: Based on the three sets of images transmitted by each camera, extract feature points and use Zhang Zhengyou's calibration method to solve for the intrinsic parameter matrices M of the first camera 21, the second camera 22, and the third camera 23 respectively. in1 M in2 M in3 , which serve as internal parameters for each camera.

[0103] Specifically, the process of calibrating the external parameters between cameras may include:

[0104] Step S14: Control the first camera 21 and the second camera 22 to acquire calibration images a1 and b1 respectively from the checkerboard target 5 of the measuring points A and B corresponding to the first cross section;

[0105] Step S15: Based on the calibration images a1 and b1, use OpenCV's findChessboardComersSB() function to extract the two-dimensional coordinates of feature points in the calibration images. Then, combine the internal parameters of each camera and substitute them into the PnP algorithm to solve for the pose transformation matrix between the target of the first cross-section corresponding to measurement points A and B and the first camera 21 and the second camera 22 under this pose.

[0106] Step S16: After controlling the gimbal 3 to change the pose of each camera, repeat steps S14 and S15.

[0107] Step S17: Repeat step S16 several times to obtain multiple sets of pose transformation matrices, thereby solving for the extrinsic parameter matrix H from the first camera 21 to the second camera 22. 12 ;

[0108] Specifically, the process of solving the extrinsic parameter matrices of the first camera 21 to the second camera 22 may include:

[0109] Step S171: Denote the coordinate systems of measuring point A, measuring point B, first camera 21, and second camera 22 as O. A O B The coordinates of any point P on the checkerboard target 5 at measuring point A, along with O1 and O2, in the above coordinate system are denoted as P1, O2 ... A P B Based on the rigid body transformation law, for each pose state (P1, P2), the following can be obtained:

[0110]

[0111] In the formula H BA Let H be the pose transformation matrix from measurement point B to measurement point A. A1 Let H be the external parameter matrix from measurement point A to the first camera 21. 2B The external parameter matrix from the second camera 22 to the measuring point B is derived as follows:

[0112] H BA H 2B H 12 H A1 =I

[0113] Step S172: Using the obtained multiple sets of pose transformation matrices, the following set of equations is obtained:

[0114]

[0115]

[0116] ...

[0117]

[0118] In the formula, the superscript (j) represents the data after the j-th pose change, 0≤j≤n, and n is the number of pose changes. Let be the external parameter matrix from the first camera to measurement point A under the initial pose. Let be the external parameter matrix from the second camera to measurement point B under the initial pose; let and Let the homogeneous transformation matrices of the motion trajectories of the first camera 21 and the second camera 22 from their initial poses to their i-th poses be represented respectively, thus obtaining:

[0119] H2 (i) H 12 =H 12 H1 (i)

[0120] Step S173: Solve the system of equations in step S172 using the hand-eye calibration method to obtain the external parameter matrices H of the first camera 21 to the second camera 22. 12 .

[0121] Specifically, the hand-eye calibration method is used to solve the equation derived in step S172. This method specifically employs a step-by-step solution method based on quaternions to solve H. 12 .

[0122] Step S18: Using the methods in steps S14 to S17, solve for the extrinsic parameter matrix H of the first camera 21 to the third camera 23. 13 ;

[0123] Step S19: Using the methods in steps S14 to S17, solve for the extrinsic parameter matrix H of the second camera 22 to the third camera 23. 23 .

[0124] In the specific implementation of step S2, the image acquisition device 2 is controlled to acquire target-containing images at measuring points A, B, and C, and the tunnel convergence deformation and arch settlement at the first cross-section are calculated using the target-containing images; this may include the following sub-steps:

[0125] Step S21: Control the first camera 21, the second camera 22 and the third camera 23 to acquire images of the checkerboard target 5 on the corresponding measurement points A, B and C of the cross section, respectively, to obtain target image a, target image b and target image c.

[0126] Step S22: Using the target image a and target image b, extract feature points, select the target world coordinate system where the measurement point A is located as the main coordinate system, and use the external parameters between the cameras to calculate the coordinates (x, y, y) of the measurement point B in the main coordinate system according to the following formula. b y b , z b ):

[0127]

[0128] In the formula M exA M is the external parameter matrix of the checkerboard target 5 from the first camera 21 to the measuring point A corresponding to the first cross-section. exB This is the external parameter matrix H of the checkerboard target 5 from the second camera 22 to the measuring point B corresponding to the first cross-section. 12 For the external parameter matrix of the first camera 21 to the second camera 22;

[0129] Step S23: Using the target image a and target image c, extract feature points, select the target world coordinate system at measurement point A as the main coordinate system, and use the external parameters between the cameras to calculate the coordinates (x, y, c) of measurement point C in the main coordinate system according to the following formula. c y c , z c ):

[0130]

[0131] Formula M exC This is the external parameter matrix of the checkerboard target 5 from the third camera 23 to the measuring point C corresponding to the first cross section at this time.

[0132] Specifically, the implementation of step S23 is the same as that of step S22, and will not be repeated here.

[0133] Step S24: Using the coordinates of measuring points B and C in the main coordinate system, calculate the length of side AB and the height h on the base AB of the survey line triangle ABC according to the following formula:

[0134]

[0135]

[0136]

[0137] p = (AB + AC + BC) / 2

[0138]

[0139] Step S25: Based on the length of side AB and the height h on the bottom side AB, calculate the tunnel convergence deformation Δd and tunnel arch settlement ΔH of the first cross-section:

[0140] Δd=AB-A′B′

[0141] ΔH=h-h'

[0142] In the formula, A′B′ is the length between measuring points A and B obtained in the previous inspection, and h′ is the height on the base AB of triangle ABC obtained in the previous inspection.

[0143] In the specific implementation of step S3, the mobile monitoring platform 1 is controlled to move to the second section, and the tunnel convergence deformation and arch settlement at the second section are calculated using the process of step S2.

[0144] Specifically, the method for determining whether the mobile detection platform has moved to the second cross-section is the same as in step S1, and will not be repeated here.

[0145] In the specific implementation of step S4, step S3 is repeated until the tunnel convergence deformation and arch settlement at all cross sections are calculated, and the mobile monitoring platform 1 is controlled to return.

[0146] As can be seen from the above embodiments, this method improves data acquisition efficiency and measurement accuracy, reduces reliance on manual labor, obtains more comprehensive cross-sectional deformation information, and can monitor the convergence of tunnel cross-sections and the settlement deformation of the arch, which has great advantages in the field of measurement.

[0147] Corresponding to the aforementioned embodiments of the tunnel deformation inspection method based on monocular pose estimation, this application also provides embodiments of a tunnel deformation inspection device based on monocular pose estimation.

[0148] Figure 6 This is a block diagram of a tunnel deformation inspection device based on monocular pose estimation, according to an exemplary embodiment. (Refer to...) Figure 6 The device may include:

[0149] The calibration module 100 is used to control the mobile monitoring platform 1 to move to the first section and use the image acquisition device 2 to calibrate the internal parameters of the first camera 21, the second camera 22 and the third camera 23 as well as the external parameters between the cameras;

[0150] The first calculation module 200 is used to control the image acquisition device 2 to acquire target-containing images at measuring points A, B, and C, and to calculate the tunnel convergence deformation and arch settlement at the first section through the target-containing images.

[0151] The second calculation module 300 is used to control the mobile monitoring platform 1 to move to the second section and calculate the tunnel convergence deformation and arch settlement at the second section using the process of the first calculation module.

[0152] Return module 400 is used to repeat the process of the second calculation module until the tunnel convergence deformation and arch settlement at all cross sections are calculated, and then control the mobile monitoring platform 1 to return.

[0153] Regarding the apparatus in the above embodiments, the specific manner in which each module performs its operation has been described in detail in the embodiments related to the method, and will not be elaborated upon here.

[0154] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this application according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0155] Accordingly, this application also provides an electronic device, including: one or more processors; a memory for storing one or more programs; when the one or more programs are executed by the one or more processors, the one or more processors implement the tunnel deformation inspection method based on monocular pose estimation as described above. Figure 7 The diagram shown is a hardware structure diagram of any device with data processing capabilities, used in an embodiment of the present invention to provide a tunnel deformation inspection method based on monocular pose estimation. (Except for...) Figure 7 In addition to the processor, memory, and network interface shown, any data processing device in the embodiment may also include other hardware depending on the actual function of the data processing device, which will not be described in detail here.

[0156] Accordingly, this application also provides a computer-readable storage medium storing computer instructions, which, when executed by a processor, implement the tunnel deformation inspection method based on monocular pose estimation as described above. The computer-readable storage medium can be an internal storage unit of any data processing device as described in any of the foregoing embodiments, such as a hard disk or memory. The computer-readable storage medium can also be an external storage device, such as a plug-in hard disk, smart media card (SMC), SD card, flash card, etc., equipped on the device. Furthermore, the computer-readable storage medium can include both internal storage units of any data processing device and external storage devices. The computer-readable storage medium is used to store the computer program and other programs and data required by the data processing device, and can also be used to temporarily store data that has been output or will be output.

[0157] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the disclosure herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein.

[0158] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope.

Claims

1. A tunnel deformation inspection method based on monocular pose estimation, characterized in that, A microcomputer terminal is used in a tunnel deformation inspection system based on monocular pose estimation. The tunnel deformation inspection system includes: Mobile monitoring platform is used to provide information on the movement of the tunnel deformation inspection system along the tunnel direction; The image acquisition device is mounted on the upper part of the mobile monitoring platform via a pan-tilt unit. It includes three cameras rigidly connected in a T-shape. The first and second cameras are used to acquire images of checkerboard targets at measuring points A and B on both sides of each section of the tunnel, respectively. The third camera is used to acquire images of checkerboard targets at measuring point C on the arch of each section of the tunnel. The checkerboard targets at measuring points A, B, and C are preset. An odometer is used to acquire mileage data after the mobile monitoring platform enters the tunnel and to determine the data collection location of the mobile monitoring platform. A microcomputer terminal is used to control the mobile detection platform to enter the tunnel; receive mileage data provided by the odometer and control the image acquisition device to acquire images based on the mileage data; and calculate the convergence deformation of each section and the settlement of the arch based on the acquired images. Tunnel deformation inspection methods include: Step S1: Control the mobile monitoring platform to move to the first section, and use the images acquired by the image acquisition device to calibrate the internal parameters of the first camera, the second camera, and the third camera, as well as the external parameters between the cameras; Step S2: Control the image acquisition device to acquire target-containing images at measuring points A, B, and C, and calculate the tunnel convergence deformation and arch settlement at the first section using the target-containing images; Step S3: Control the mobile monitoring platform to move to the second section, and calculate the tunnel convergence deformation and arch settlement at the second section using the process in step S2; Step S4: Repeat step S3 until the tunnel convergence deformation and arch settlement at all cross sections have been calculated, and then control the mobile monitoring platform to return; Step S2 includes: Step S21: Control the first camera, the second camera and the third camera to acquire images of the checkerboard target on the measurement point A, measurement point B and measurement point C corresponding to a certain cross section, respectively, to obtain target image a, target image b and target image c; Step S22: Using the target image a and target image b, extract feature points, select the target world coordinate system where the measurement point A is located as the main coordinate system, and use the external parameters between the cameras to calculate the coordinates of the measurement point B in the main coordinate system according to the following formula ( ): ， In the formula This is the external parameter matrix of the checkerboard target from the first camera to the measuring point A corresponding to the first cross-section. This is the external parameter matrix of the checkerboard target from the second camera to the measuring point B corresponding to the first cross-section. This is the external parameter matrix from the first camera to the second camera; Step S23: Using the target image a and target image c, extract feature points, select the target world coordinate system at measurement point A as the main coordinate system, and use the external parameters between the cameras to calculate the coordinates of measurement point C in the main coordinate system according to the following formula ( ): ， Mode This is the external parameter matrix of the checkerboard target from the third camera to the measuring point C corresponding to the first cross section at this time; Step S24: Using the coordinates of measuring points B and C in the main coordinate system, calculate the length of side AB and the height h on the base AB of the survey line triangle ABC according to the following formula: ， ， ， ， ， Step S25: Based on the length of side AB and the height h on the bottom side AB, calculate the tunnel convergence deformation Δd and tunnel arch settlement ΔH of the first cross-section: = – ， ， In the formula This is the length between measuring points A and B, as measured in the last inspection. This is the altitude on the base AB of triangle ABC obtained from the previous inspection.

2. The method according to claim 1, characterized in that, In step S1, controlling the mobile monitoring platform to move to the first cross-section and using the images acquired by the image acquisition device to calibrate the internal parameters of the first camera, the second camera, and the third camera includes: Step S11: Control the movement of the mobile monitoring platform according to the real-time mileage data transmitted by the odometer. When the mileage data transmitted by the odometer matches the mileage information of the first section to be measured, control the mobile monitoring platform to stop moving. At this time, the mobile monitoring platform reaches the first section. Step S12: Control the first camera, the second camera and the third camera to acquire several images of the target at measurement points A, B and C corresponding to the cross section respectively; Step S13: Based on the three sets of images transmitted by each camera, extract feature points and use Zhang Zhengyou's calibration method to solve for the intrinsic parameter matrices of the first camera, the second camera, and the third camera respectively. , , , which serve as internal parameters for each camera.

3. The method according to claim 1, characterized in that, In step S1, calibrating the external parameters between cameras includes: Step S14: Control the first camera and the second camera to acquire calibration image a1 and calibration image b1 respectively from the checkerboard target of measuring point A and measuring point B corresponding to the first cross section; Step S15: Based on the calibration images a1 and b1, extract feature points. Using the internal parameters of each camera, employ the PnP algorithm to solve for the pose transformation matrix between the target at measurement points A and B corresponding to the first cross-section and the first and second cameras, respectively, under this pose. , ; Step S16: After controlling the gimbal to change the pose of each camera, repeat steps S14 and S15. Step S17: Repeat step S16 several times to obtain multiple sets of pose transformation matrices, thereby solving for the extrinsic parameter matrix from the first camera to the second camera. ; Step S18: Using steps S14~S17, solve for the extrinsic parameter matrix from the first camera to the third camera. ; Step S19: Using steps S14~S17, solve for the extrinsic parameter matrix from the second camera to the third camera. .

4. The method according to claim 3, characterized in that, Solving for the extrinsic parameter matrix from the first camera to the second camera includes: Step S171: Denote the coordinate systems of measuring point A, measuring point B, and the first and second cameras as follows: , and , The coordinates of any point P on the checkerboard target at measuring point A in the coordinate system are denoted as follows: , , , Based on the rigid body transformation law, the following can be obtained for each pose state: ， In the formula Let be the pose transformation matrix from measurement point B to measurement point A. Let A be the external parameter matrix from measurement point A to the first camera. Given the extrinsic parameter matrix from the second camera to measurement point B, we can derive: ， Step S172: Using the obtained multiple sets of pose transformation matrices, the following set of equations is obtained: ， ， …… ， superscript ( j ) represents the first j Data after the next pose change, 0≤ j ≤n, where n is the number of pose changes. Let be the external parameter matrix from the first camera to measurement point A under the initial pose. Let be the external parameter matrix from the second camera to measurement point B under the initial pose; let and Representing the first and second cameras from their initial poses to the... The homogeneous transformation matrix of each pose motion trajectory is obtained, thus: ， Step S173: Solve the system of equations in step S172 to obtain the extrinsic parameter matrix from the first camera to the second camera. .

5. A tunnel deformation inspection device based on the tunnel deformation inspection method based on monocular pose estimation as described in claim 1, characterized in that, This is achieved through a microcomputer terminal in the tunnel deformation inspection system, including: The calibration module is used to control the mobile monitoring platform to move to the first cross section and to calibrate the internal parameters of the first camera, the second camera and the third camera, as well as the external parameters between the cameras, using the images acquired by the image acquisition device. The first calculation module is used to control the image acquisition device to acquire target-containing images at measuring points A, B, and C, and to calculate the tunnel convergence deformation and arch settlement at the first cross section through the target-containing images. The second calculation module is used to control the mobile monitoring platform to move to the second section and calculate the tunnel convergence deformation and arch settlement at the second section using the process of the first calculation module. The return module is used to repeat the process of the second calculation module until the tunnel convergence deformation and arch settlement at all cross sections are calculated, and then control the mobile monitoring platform to return.

6. An electronic device, characterized in that, include: One or more processors; Memory, used to store one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in any one of claims 1-4.

7. A computer-readable storage medium storing computer instructions thereon, characterized in that, When executed by the processor, this instruction implements the steps of the method as described in any one of claims 1-4.

Images