Tunnel disease data extraction and calibration method

Abstract

The invention discloses a tunnel disease data extraction and calibration method, and particularly relates to the technical field of tunnel disease detection. Comprising the following steps: S1, establishing a tunnel BIM model coordinate system; s2, calibrating the positioning precision of a high-definition camera and a three-dimensional laser scanner through a tunnel BIM model coordinate system; s3, collecting tunnel disease information by using a high-definition camera and a three-dimensional laser scanner; s4, extracting feature points of the tunnel disease information; and S5, converting the two-dimensional coordinates of the feature points into three-dimensional coordinates in an actual space. By adopting the technical scheme of the invention, the problem of low positioning accuracy in a traditional tunnel disease detection method is solved, the accuracy and efficiency of tunnel disease detection are improved, and the safety and stability of a tunnel structure are ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of tunnel disease detection, and in particular to a method for extracting and calibrating tunnel disease data. Background Art

[0002] As an important part of transportation infrastructure, the structural safety and stability of tunnels are crucial to the smooth operation of transportation. During the long-term use of tunnels, cracks, leakages, spalling and other defects are prone to occur due to factors such as natural environment, geological changes and construction quality, which affect the safety and life of the tunnel. Therefore, it is particularly important to conduct timely and accurate detection and maintenance of tunnel defects.

[0003] Traditional tunnel disease detection methods mainly rely on manual inspections, which have problems such as low detection accuracy, low efficiency and strong subjectivity. With the development of science and technology, tunnel detection systems based on 3D laser scanning and high-definition camera technology are gradually being applied in engineering practice. However, the existing technology still has the following deficiencies in the detailed collection and data fusion of disease data:

[0004] (1) Insufficient positioning accuracy: Most existing tunnel disease detection systems rely on traditional positioning technologies such as GPS. In the tunnel environment, GPS signals are blocked, resulting in reduced positioning accuracy and reliability.

[0005] (2) Incomplete data collection: Single 3D laser scanning or high-definition camera technology cannot simultaneously obtain the geometric shape and detailed texture information inside the tunnel, making it difficult to fully reflect the actual situation of the disease.

[0006] (3) Difficulty in data fusion: The existing defect detection system lacks effective data fusion methods, making it difficult to accurately integrate key defect information into the tunnel 3D model, which affects subsequent analysis and maintenance work. Summary of the invention

[0007] The present invention aims to provide a tunnel disease data extraction and calibration method, which solves the problem of low positioning accuracy in traditional tunnel disease detection methods.

[0008] In order to achieve the above object, the technical solution of the present invention is as follows: a method for extracting and calibrating tunnel disease data, comprising the following steps:

[0009] S1. Establish the tunnel BIM model coordinate system;

[0010] S2. Calibrate the positioning accuracy of the high-definition camera and 3D laser scanner through the tunnel BIM model coordinate system;

[0011] S3, using high-definition cameras and 3D laser scanners to collect tunnel disease information;

[0012] S4, extracting characteristic points of tunnel disease information;

[0013] S5. Convert the two-dimensional coordinates of the feature points into three-dimensional coordinates in the actual space.

[0014] Furthermore, the coordinate system establishment method of step S1 is: at the tunnel entrance or a specific location, an obvious and easily identifiable marking point is selected as the coordinate origin to establish the global coordinate system of the tunnel BIM model.

[0015] Furthermore, the calibration method of step S2 is as follows:

[0016] S2.1. Record the coordinates of the marked points in the BIM model as reference points;

[0017] S2.2. Use the level and gyroscope to reset the level, inclination and mileage of the high-definition camera and 3D laser scanner to zero;

[0018] S2.3, start the high-definition camera and the three-dimensional laser scanner, and obtain the attitude information of the high-definition camera and the three-dimensional laser scanner in real time through the inertial measurement unit; at the same time, collect the marking point data inside the tunnel, and record the local coordinate system information of the marking point;

[0019] S2.3, processing the collected images and point cloud data, extracting the marking points on the tunnel surface, and matching the extracted marking points with the marking points in the BIM model;

[0020] S2.4, converting the two-dimensional pixel coordinates into points in three-dimensional space through the internal parameter matrix, and converting the points in the camera coordinate system into the global coordinate system through the external parameter matrix, calculating the error between the converted marker points and the marker points in the BIM model, and obtaining the position and posture deviation of the high-definition camera and the three-dimensional laser scanner;

[0021] S2.5. According to the error calculation results, adjust the positioning parameters of the high-definition camera and the three-dimensional laser scanner, and correct the position and posture of the high-definition camera and the three-dimensional laser scanner.

[0022] Furthermore, in step S2.5, the correction effect is monitored in real time through a closed-loop control system to ensure the positioning accuracy of the high-definition camera and the three-dimensional laser scanner.

[0023] Furthermore, the collected tunnel disease information is re-encoded in a sequential manner of time information, position information, spatial posture information and depth information.

[0024] Furthermore, the feature point extraction method of step S4 is: preprocessing the tunnel disease image collected in step S3, extracting the feature points of the preprocessed tunnel disease using an image processing algorithm, and extracting three-dimensional feature points corresponding to the image feature points from the point cloud data of the three-dimensional laser scanner.

[0025] Furthermore, it is characterized in that: the conversion method of step S5 is as follows:

[0026] S5.1, using the camera's internal parameter matrix to convert the two-dimensional pixel coordinates into three-dimensional points in the camera coordinate system;

[0027] S5.2. Use the data of the 3D laser scanner to obtain the depth information of the disease feature points in the 3D space, align the disease image with the depth information, and determine the 3D distance corresponding to each pixel point;

[0028] S5.3, obtain the external parameter matrix of the camera, and construct a transformation matrix from the camera to the world coordinate system;

[0029] S5.4. Insert the converted defect feature points into the tunnel BIM model to realize the three-dimensional visualization expression of the defect in the tunnel model.

[0030] Furthermore, the method of step S5.1 is as follows: Assume that the pixel coordinates of a certain point in the disease image are (u, v), and according to the internal parameter matrix of the camera, it can be converted into the coordinates of the normalized plane (x, y, 1), and the formula is:

[0031]

[0032] Among them, c x and c y is the optical center coordinate, f x and f y is the focal length.

[0033] Furthermore, in step S2, the pixel coordinates are converted into three-dimensional points in the coordinate system of the identification camera in combination with the depth information as follows: Assuming that the depth information provided by the laser scanner is Z, the coordinates of the three-dimensional point in the coordinate system of the identification camera are:

[0034] (X,Y,Z)=(x·Z,y·Z,Z).

[0035] Furthermore, in step S5.3, the method of converting the point in the camera coordinate system to the world coordinate system is as follows:

[0036] Let the three-dimensional point in the camera coordinate system be P c =(X,Y,Z), then point P in the world coordinate system w Converted by the following formula: w=R·P c +T

[0037] Where R is the rotation matrix and T is the translation vector.

[0038] Compared with the prior art, this solution has the following beneficial effects:

[0039] 1. This solution provides a defect information coding method and a positioning and calibration function based on the tunnel BIM model, combined with portable and low-cost inspection equipment to achieve efficient data fusion, thereby improving the positioning accuracy of defect inspection equipment in complex environments and the accuracy of defect detection, meeting the needs of rapid response and comprehensive display, and improving the efficiency and reliability of tunnel detection.

[0040] 2. This solution is based on the positioning and calibration function of the tunnel BIM model. By using the tunnel BIM model as the calibration reference and combining it with the markers detected by the inspection equipment for verification, it achieves precise positioning of information such as tunnels and defects, and improves the accuracy and reliability of positioning.

[0041] 3. This solution uses standardized coding of detection data: By formulating coding rules and combining information such as [position coding]-[posture coding]-[depth information coding], the specific location of the disease in the disease image can be determined through the coding content and coordinate conversion relationship. By standardizing the coding of the collected data, it can serve as a key basis for extracting disease information and determining spatial coordinates. Data coding can maximize its application value while saving storage.

[0042] 4. This solution uses 3D laser scanners and high-definition cameras to capture image data and align the two to obtain 3D feature points of the disease, making full use of the advantages of multi-source data. Compared with the detection method that only relies on schematic display, this solution can provide more detailed and accurate 3D disease models and image information, improving the accuracy and comprehensiveness of disease detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 It is a flow chart of a method for extracting and calibrating tunnel disease data according to the present invention;

[0044] Figure 2 is a schematic diagram of depth information extraction in this embodiment;

[0045] Figure 3 Schematic diagram of disease feature point coordinate extraction in this embodiment;

[0046] Figure 4 It is a schematic diagram of three-dimensional space coordinate transformation in this embodiment. DETAILED DESCRIPTION

[0047] The present invention is further described in detail below through specific embodiments:

[0048] Example

[0049] like Figure 1 As shown, a tunnel disease data extraction and calibration method comprises the following steps:

[0050] S1. Establish a tunnel BIM model coordinate system. The coordinate system establishment method is: at the tunnel entrance or a specific location, select an obvious and easily identifiable mark point as the coordinate origin to establish the global coordinate system of the tunnel BIM model.

[0051] S2. Calibrate the positioning accuracy of the high-definition camera and 3D laser scanner using the tunnel BIM model coordinate system. The calibration method is as follows:

[0052] S2.1. Record the coordinates of the marked points in the BIM model as reference points, which are used for subsequent calibration and comparison.

[0053] S2.2. Use the level and gyroscope to reset the level, inclination and mileage of the HD camera and 3D laser scanner to zero, ensuring that each sensor of the HD camera and 3D laser scanner is in the initial calibration state.

[0054] S2.3. Start the high-definition camera and the three-dimensional laser scanner, and obtain the attitude information (such as inclination, direction, etc.) of the high-definition camera and the three-dimensional laser scanner in real time through the inertial measurement unit (IMU); at the same time, collect the marking point data inside the tunnel (such as marking body, pipe seam, etc.), and record the local coordinate system information of the marking point.

[0055] S2.3. Process the collected images and point cloud data, extract the marking points on the tunnel surface, use feature point matching algorithms (such as SIFT, SURF, ORB, etc.) to match the extracted marking points with the marking points in the BIM model, eliminate incorrectly matched marking points, and ensure the accuracy of the matching results.

[0056] S2.4. Convert the two-dimensional pixel coordinates into points in three-dimensional space through the internal parameter matrix (focal length, optical center coordinates), and convert the points in the marker camera coordinate system into the global coordinate system through the external parameter matrix (position and direction of the high-definition camera and the three-dimensional laser scanner in the tunnel coordinate system), calculate the error between the converted marker points and the marker points in the BIM model, and obtain the position and posture deviation of the high-definition camera and the three-dimensional laser scanner;

[0057] S2.5. According to the error calculation results, adjust the positioning parameters (such as displacement, angle, etc.) of the HD camera and 3D laser scanner to correct the position and posture of the HD camera and 3D laser scanner. At the same time, monitor the correction effect in real time through the closed-loop control system to ensure the positioning accuracy of the HD camera and 3D laser scanner. By continuously updating the calibration parameters of the HD camera and 3D laser scanner, the positioning accuracy of the HD camera and 3D laser scanner in the entire tunnel inspection process is ensured.

[0058] S2.6. The real-time positioning status and error information of the high-definition camera and 3D laser scanner are displayed on the user interface. The operator can make real-time corrections to the positioning parameters based on the error information to improve the positioning accuracy and stability.

[0059] S3. Start a high-definition camera to shoot the tunnel lining surface section by section to obtain high-resolution disease images. Use a 3D laser scanner to synchronously scan the tunnel lining surface to obtain 3D point cloud data to collect depth data.

[0060] In this step, the collected tunnel disease information is re-encoded in the order of time information, location information, spatial posture information and depth information:

[0061] Time information: It is used to record the specific time of data collection. During the data collection process, the device will automatically generate time information. Time information is usually provided by the internal clock of the device and can be in a standard date and time format (for example: YYYY-MM-DDTHH:MM:SS). Time information helps track the generation time and timeliness of data in subsequent data processing and analysis.

[0062] Position information: It is used to identify the specific position of the camera in three-dimensional space. The three-dimensional coordinates of the camera are obtained through an inertial measurement unit (IMU) or other positioning technology. Position information includes the specific coordinates of the camera in the world coordinate system, which is usually recorded in the form of coordinate values ​​(such as X, Y, Z). This can provide a spatial reference for the disease image so that the acquisition position of each image can be accurately determined.

[0063] Spatial attitude information: describes the spatial attitude information of the identification camera, including the tilt angle and direction. The attitude data of the identification camera is obtained through the inertial measurement unit (IMU), which includes the pitch, yaw and roll angles of the camera. The attitude information is usually recorded as an angle value and combined with the positioning data of the identification camera to accurately interpret the shooting angle and direction of the image, thereby ensuring the accuracy and consistency of the disease information.

[0064] Depth information: The grid point cloud data storage method is used to record the depth information at the grid intersection to reduce the amount of data storage, and the interpolation method is used in subsequent processing to infer the depth information of the internal points of the grid to infer the missing depth information. Figure 2 As shown in the figure, a 3D laser scanner is used to scan the tunnel surface to obtain a series of depth data. The gridding method only retains the depth information at the grid intersection, that is, only the actual depth value of the grid node is recorded. The data storage format is usually an array of depth values ​​of the grid nodes. The depth information of the points inside the grid is inferred by the interpolation algorithm. This method not only reduces the data storage requirements, but also ensures the accuracy of the point cloud data as much as possible, which is convenient for subsequent spatial analysis and 3D reconstruction.

[0065] By standardizing the encoding of the collected data, it can serve as the key basis for disease information extraction and spatial coordinate determination. Data encoding can save storage while maximizing its application value.

[0066] S4, such as Figure 3 As shown, the captured defect image is preprocessed, including denoising, contrast enhancement, etc. Then, the feature points of tunnel defect information are extracted using image processing algorithms (such as Canny edge detection, Harris corner detection, etc.). From the point cloud data of the 3D laser scanner, 3D feature points corresponding to the image feature points are extracted.

[0067] S5, such as Figure 4 As shown, the two-dimensional coordinates of the feature points are converted into three-dimensional coordinates in the actual space. The conversion method is as follows:

[0068] S5.1. Use the camera's internal parameter matrix (including focal length and optical center coordinates) to convert the two-dimensional pixel coordinates into three-dimensional points in the camera coordinate system. The method is as follows: Let the pixel coordinates of a point in the disease image be (u, v). According to the camera's internal parameter matrix, it can be converted into the coordinates of the normalized plane (x, y, 1). The formula is:

[0069]

[0070] Among them, c x and c y is the optical center coordinate, f x and f y is the focal length.

[0071] S5.2. Use the data from the 3D laser scanner to obtain the depth information of the disease feature points in the 3D space, align the disease image with the depth information, and determine the 3D distance corresponding to each pixel point; combine the depth information to convert the pixel coordinates into 3D points in the camera coordinate system:

[0072] Assuming that the depth information provided by the laser scanner is Z, the coordinates of the 3D point in the camera coordinate system are:

[0073] (X,Y,Z)=(x·Z,y·Z,Z).

[0074] S5.3. Obtain the external parameter matrix of the high-definition camera (including the position and orientation of the camera in the tunnel coordinate system), and construct a transformation matrix that identifies the camera to the world coordinate system; wherein the external parameter matrix includes a rotation matrix R and a translation vector T, which represent the transformation of the camera coordinate system relative to the world coordinate system.

[0075] Use the external parameter matrix to transform the point in the camera coordinate system to the world coordinate system:

[0076] Let the three-dimensional point in the camera coordinate system be P c =(X,Y,Z), then point P in the world coordinate system w Converted by the following formula: w =R·P c +T

[0077] S5.4. Insert the converted defect feature points into the tunnel BIM model to realize the three-dimensional visualization of the defect in the tunnel model. Compare the positions of the defect feature points in the BIM model with the actual detected positions, perform error analysis and calibration, and ensure the accuracy of the data.

[0078] The above are only embodiments of the present invention, and the common knowledge such as the known specific structures and / or characteristics in the scheme are not described in detail here. It should be pointed out that for those skilled in the art, several deformations and improvements can be made without departing from the structure of the present invention, which should also be regarded as the protection scope of the present invention, and these will not affect the effect of the implementation of the present invention and the practicality of the patent. The scope of protection required by this application shall be based on the content of its claims, and the specific implementation methods and other records in the specification can be used to interpret the content of the claims.

Claims

1. A method for extracting and calibrating tunnel disease data, characterized in that: The steps include: S1. Establish the tunnel BIM model coordinate system; S2. Calibrate the positioning accuracy of the high-definition camera and 3D laser scanner through the tunnel BIM model coordinate system; S3, using high-definition cameras and 3D laser scanners to collect tunnel disease information; S4, extracting characteristic points of tunnel disease information; S5. Convert the two-dimensional coordinates of the feature points into three-dimensional coordinates in the actual space.

2. A method for extracting and calibrating tunnel disease data according to claim 1, characterized in that: The method for establishing the coordinate system in step S1 is as follows: at the tunnel entrance or a specific location, an obvious and easily identifiable marking point is selected as the coordinate origin to establish the global coordinate system of the tunnel BIM model.

3. A method for extracting and calibrating tunnel disease data according to claim 2, characterized in that: The calibration method of step S2 is as follows: S2.

1. Record the coordinates of the marked points in the BIM model as reference points; S2.

2. Use the level and gyroscope to reset the level, inclination and mileage of the high-definition camera and 3D laser scanner to zero; S2.3, start the high-definition camera and the three-dimensional laser scanner, and obtain the attitude information of the high-definition camera and the three-dimensional laser scanner in real time through the inertial measurement unit; at the same time, collect the marking point data inside the tunnel, and record the local coordinate system information of the marking point; S2.3, processing the collected images and point cloud data, extracting the marking points on the tunnel surface, and matching the extracted marking points with the marking points in the BIM model; S2.4, converting the two-dimensional pixel coordinates into points in three-dimensional space through the internal parameter matrix, and converting the points in the camera coordinate system into the global coordinate system through the external parameter matrix, calculating the error between the converted marker points and the marker points in the BIM model, and obtaining the position and posture deviation of the high-definition camera and the three-dimensional laser scanner; S2.

5. According to the error calculation results, adjust the positioning parameters of the high-definition camera and the three-dimensional laser scanner, and correct the position and posture of the high-definition camera and the three-dimensional laser scanner.

4. A method for extracting and calibrating tunnel disease data according to claim 3, characterized in that: In step S2.5, the correction effect is monitored in real time through a closed-loop control system to ensure the positioning accuracy of the high-definition camera and the three-dimensional laser scanner.

5. The method for extracting and calibrating tunnel disease data according to claim 1, characterized in that: The collected tunnel disease information is re-encoded in the order of time information, location information, spatial posture information and depth information.

6. The method for extracting and calibrating tunnel disease data according to claim 1, characterized in that: The feature point extraction method of step S4 is: preprocessing the tunnel disease image collected in step S3, extracting the feature points of the preprocessed tunnel disease using an image processing algorithm, and extracting three-dimensional feature points corresponding to the image feature points from the point cloud data of the three-dimensional laser scanner.

7. A method for extracting and calibrating tunnel disease data according to claim 6, characterized in that: The conversion method of step S5 is as follows: S5.1, using the camera's internal parameter matrix to convert the two-dimensional pixel coordinates into three-dimensional points in the camera coordinate system; S5.

2. Use the data of the 3D laser scanner to obtain the depth information of the disease feature points in the 3D space, align the disease image with the depth information, and determine the 3D distance corresponding to each pixel point; S5.3, obtain the external parameter matrix of the camera, and construct a transformation matrix from the camera to the world coordinate system; S5.

4. Insert the converted defect feature points into the tunnel BIM model to realize the three-dimensional visualization expression of the defect in the tunnel model.

8. A method for extracting and calibrating tunnel disease data according to claim 7, characterized in that: The method of step S5.1 is as follows: Assume that the pixel coordinates of a point in the disease image are (u, v). According to the internal parameter matrix of the camera, it can be converted into the coordinates of the normalized plane (x, y, 1). The formula is: Among them, c x and c y is the optical center coordinate, f x and f y is the focal length.

9. The method for extracting and calibrating tunnel disease data according to claim 7, characterized in that: In step S2, the pixel coordinates are converted into three-dimensional points in the identification camera coordinate system in combination with the depth information as follows: Assuming that the depth information provided by the laser scanner is Z, the coordinates of the three-dimensional point in the identification camera coordinate system are: (X,Y,Z)=(x·Z,y·Z,Z).

10. The method for extracting and calibrating tunnel disease data according to claim 7, characterized in that: In step S5.3, the method of converting the point in the camera coordinate system to the world coordinate system is as follows: Let the three-dimensional point in the camera coordinate system be P c =(X,Y,Z), then point P in the world coordinate system w Converted by the following formula: w =R·P c +T Where R is the rotation matrix and T is the translation vector.

Images