A tunnel surface deformation monitoring method based on multi-baseline close-range photogrammetry

Abstract

The present invention discloses a tunnel surface deformation monitoring method based on multi-baseline close-range photogrammetry, comprising the following steps: S1: original image, using a CCD camera installed at a position in the tunnel that needs to be monitored, to collect its original image and store it; S2: reconstruction of the original image three-dimensional point cloud model, to reconstruct the three-dimensional model of the collected original image; S3: real-time image, using the CCD camera installed at the position in the tunnel that needs to be monitored in S1 to take regular photos of the position of the tunnel and store it; S4: reconstruction of the real-time image three-dimensional point cloud model. The present invention effectively realizes real-time monitoring of the use of the tunnel by monitoring the difference between the real-time image and the original image, can well control the service life and safety of the tunnel, and effectively reduce the losses caused by the use of the tunnel. At the same time, the method can realize real-time unmanned monitoring, and its monitoring is reliable.

Description

Technical Field

[0001] The invention relates to the technical field of close-range photogrammetry, and in particular to a tunnel surface deformation monitoring method based on multi-baseline close-range photogrammetry. Background Art

[0002] Deformation monitoring of municipal, bridge, water conservancy, civil engineering and other projects is an important part of structural health monitoring and an important indicator for evaluating structural stability. Conventional measurement methods represented by level and total station are labor-intensive and greatly affected by the operation mode of the instrument. Therefore, deformation monitoring and measurement technology based on machine vision has been derived, which integrates photogrammetry, image processing and computer technology. By processing the image by computer and comparing the changes of the target point image in the image sequence, the two-dimensional displacement deformation can be calculated. For example, the target is installed on the target structure, and a monocular camera can be used to effectively monitor the displacement of the target in the plane parallel to its imaging surface in a two-dimensional plane. However, the monocular camera does not have a ranging function. When the target is deflected due to external force or the distance between the target and the camera changes, the displacement of the target in the two-dimensional plane cannot be accurately identified by image recognition technology. In addition, because the distance between the target and the camera changes, the calibration size of the original image pixel point will be inaccurate, so there will be a large error in the measurement of the target displacement. Therefore, a tunnel surface deformation monitoring method based on multi-baseline close-range photogrammetry is urgently needed to solve the above problems. Summary of the invention

[0003] Based on the technical problems existing in the background technology, the present invention proposes a tunnel surface deformation monitoring method based on multi-baseline close-range photogrammetry.

[0004] The present invention proposes a tunnel surface deformation monitoring method based on multi-baseline close-range photogrammetry, comprising the following steps:

[0005] S1: Original image, using a CCD camera installed at the location of the tunnel that needs to be monitored to collect and store the original image;

[0006] S2: reconstruction of the three-dimensional point cloud model of the original image, reconstruction of the three-dimensional model of the collected original image;

[0007] S3: real-time image, using the CCD camera installed at the position of the tunnel to be monitored in S1 to take regular photos of the position of the tunnel and store them;

[0008] S4: reconstruction of the three-dimensional point cloud model of the real-time image, reconstruction of the three-dimensional model of the collected real-time image;

[0009] S5: Perform 3D model similarity topology comparison on the 3D point cloud reconstruction models in S2 and S4;

[0010] S6: The difference between the early and late deformations of the tunnel is obtained through S5, so as to estimate the damage of the tunnel during use.

[0011] Furthermore, in S1 and S3, the original image is collected and stored by using the light from the outside or the light reflected by the object to irradiate the CCD image sensor through the lens to obtain an electrical signal, and then the image signal processing circuit and the amplification circuit are used to generate a digital image of each row or multiple rows in coordination with the clock synchronization signal. The final image is stored on the memory card through the digital image, which is converted into a two-dimensional digital signal and a black and white image is generated. A layer of red, green and blue film is covered on the sensor, and any color can be expressed through the three primary colors to simulate a color, so the color sensor generates color information. Add another layer of film on top, and use the photolithography method to carve out a tiny lens to ensure that the light can converge on the small grid at the bottom through the lens, which can improve the sensor's sensitivity to light and conversion efficiency.

[0012] Furthermore, in S2 and S4, the reconstruction of the original image three-dimensional point cloud model includes the following steps:

[0013] Step 1: Integrate the spatial point cloud data and color intensity data acquired by the RGB-D camera and manage and output them in an engineering manner;

[0014] Step 2: Use PCL technology (a modular, cross-platform, open source C++ programming library for 3D point cloud processing) to pre-process the point cloud data, including denoising, segmentation, filtering, registration, sampling and other operations, and output a point cloud with obvious features and streamlined data.

[0015] Step 3: Meshing the point cloud data, that is, using a series of grids to approximate the point cloud, generally using triangular grids and quadrilateral grids. This step realizes the conversion of point cloud to mesh in three-dimensional representation.

[0016] Step 4: Panoramic texture mapping: Map the color and texture information collected by the RGB-D camera onto the mesh model, refine it, and output a realistic 3D model.

[0017] Furthermore, in S3, when the CCD camera installed at the position of the tunnel to be monitored takes timed photos of the position of the tunnel, the time of the timed photos can be set by people, specifically, 2 minutes, 5 minutes, 8 minutes and 10 minutes.

[0018] Furthermore, in S5, each three-dimensional value in the original image is defined as a target Oi...On, and each three-dimensional value of a feature point in the real-time image is defined as Di...Dn.

[0019] Furthermore, the three-dimensional coordinates (Xd, Yd, Nd) of the feature point of each real-time image are compared with the three-dimensional coordinates (Xi, Yi, Ni).

[0020] Furthermore, if the values ​​of the two are all equal, it means that the three-dimensional point cloud models of the original image and the real-time image have not changed, so the deformation of the next frame of the tunnel can continue to be monitored.

[0021] Furthermore, if the values ​​of the two are not equal, for example, Xd and Xi are equal, but Yd and Yi are not equal to Nd and Ni, the values ​​of the two are not equal. Similarly, if two values ​​are equal and one value is not equal, it can also be defined as unequal, which means that the three-dimensional point cloud model of the original image and the real-time image has changed, so that the deformation of the tunnel can be judged. If the unequal values ​​are within the controllable range, the CCD camera can continue to monitor the deformation of the next frame of the tunnel.

[0022] Furthermore, if the values ​​of the two are not equal, for example, Xd and Xi are equal, but Yd and Yi are not equal to Nd and Ni, that is, the values ​​of the two are not equal. Similarly, if two values ​​are equal and one value is not equal, it can also be defined as unequal, which means that the three-dimensional point cloud model of the original image and the real-time image has changed, so that the deformation of the tunnel can be judged. If the unequal values ​​are not within the set range, the deformation of the tunnel cannot maintain the continued use of the tunnel, and the deformed position of the tunnel can be repaired and abandoned.

[0023] The present invention has the following beneficial effects:

[0024] By monitoring the difference between the real-time image and the original image, the real-time monitoring of the tunnel usage can be effectively realized, which can well control the service life and safety of the tunnel and effectively reduce the losses caused by the use of the tunnel. At the same time, this method can realize real-time unmanned monitoring, and its monitoring is reliable. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 It is a schematic diagram of a tunnel surface deformation monitoring method based on multi-baseline close-range photogrammetry according to the present invention. DETAILED DESCRIPTION

[0026] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0027] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art of the present invention. The terms used herein in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The term "and / or" used herein includes any and all combinations of one or more related listed items.

[0028] Reference Figure 1 , a tunnel surface deformation monitoring method based on multi-baseline close-range photogrammetry, comprising the following steps:

[0029] S1: Original image, using a CCD camera installed at the location of the tunnel that needs to be monitored to collect and store the original image;

[0030] S2: reconstruction of the three-dimensional point cloud model of the original image, reconstruction of the three-dimensional model of the collected original image;

[0031] S3: real-time image, using the CCD camera installed at the position of the tunnel to be monitored in S1 to take regular photos of the position of the tunnel and store them;

[0032] S4: reconstruction of the three-dimensional point cloud model of the real-time image, reconstruction of the three-dimensional model of the collected real-time image;

[0033] S5: Perform 3D model similarity topology comparison on the 3D point cloud reconstruction models in S2 and S4;

[0034] S6: The difference between the early and late deformations of the tunnel is obtained through S5, so as to estimate the damage of the tunnel during use.

[0035] Furthermore, in S1 and S3, the original image is collected and stored by using the light from the outside or the light reflected by the object to irradiate the CCD image sensor through the lens to obtain an electrical signal, and then the image signal processing circuit and the amplification circuit are used to generate a digital image of each row or multiple rows in coordination with the clock synchronization signal. The final image is stored on the memory card through the digital image, which is converted into a two-dimensional digital signal and a black and white image is generated. A layer of red, green and blue film is covered on the sensor, and any color can be expressed through the three primary colors to simulate a color, so the color sensor generates color information. Add another layer of film on top, and use the photolithography method to carve out a tiny lens to ensure that the light can converge on the small grid at the bottom through the lens, which can improve the sensor's sensitivity to light and conversion efficiency.

[0036] Furthermore, in S2 and S4, the reconstruction of the original image three-dimensional point cloud model includes the following steps:

[0037] Step 1: Integrate the spatial point cloud data and color intensity data acquired by the RGB-D camera and manage and output them in an engineering manner;

[0038] Step 2: Use PCL technology (a modular, cross-platform, open source C++ programming library for 3D point cloud processing) to pre-process the point cloud data, including denoising, segmentation, filtering, registration, sampling and other operations, and output a point cloud with obvious features and streamlined data.

[0039] Step 3: Meshing the point cloud data, that is, using a series of grids to approximate the point cloud, generally using triangular grids and quadrilateral grids. This step realizes the conversion of point cloud to mesh in three-dimensional representation.

[0040] Step 4: Panoramic texture mapping: Map the color and texture information collected by the RGB-D camera onto the mesh model, refine it, and output a realistic 3D model.

[0041] Furthermore, in S3, when the CCD camera installed at the position to be monitored in the tunnel takes timed photos of the position of the tunnel, the time of the timed photos can be set by people, specifically 2 minutes, 5 minutes, 8 minutes and 10 minutes. There are multiple options for setting the interval time here, and people can make specific settings according to the usage of the tunnel.

[0042] Furthermore, in S5, each three-dimensional numerical value in the original image is defined as a target Oi...On, and the three-dimensional numerical value of each feature point in the real-time image is defined as Di...Dn; and the three-dimensional coordinates (Xd, Yd, Nd) of the feature point of each real-time image are compared with the three-dimensional coordinates (Xi, Yi, Ni).

[0043] By comparing the above comparison values, the following conclusions are drawn:

[0044] 1. If the values ​​of the two are all equal, it means that the 3D point cloud model of the original image and the real-time image has not changed, so the deformation of the next frame of the tunnel can continue to be monitored.

[0045] 2. If the values ​​of the two are not equal, for example, Xd and Xi are equal, but Yd and Yi are not equal to Nd and Ni, the values ​​of the two are not equal. Similarly, if two values ​​are equal and one value is not equal, it can also be defined as unequal. This means that the three-dimensional point cloud model of the original image and the real-time image has changed, so that the deformation of the tunnel can be judged. If the unequal values ​​are within the controllable range, the CCD camera can continue to monitor the deformation of the next frame of the tunnel.

[0046] 3. If the values ​​of the two are not equal, for example, Xd and Xi are equal, but Yd and Yi are not equal to Nd and Ni, the values ​​of the two are not equal. Similarly, if two values ​​are equal and one value is not equal, it can also be defined as unequal. This means that the three-dimensional point cloud model of the original image and the real-time image has changed, so the deformation of the tunnel can be judged. If the unequal values ​​are not within the set range, the deformation of the tunnel cannot maintain the continued use of the tunnel, and the deformed position of the tunnel can be repaired and abandoned.

[0047] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A tunnel surface deformation monitoring method based on multi-baseline close-range photogrammetry, characterized in that: The following steps are involved: S1: Original image, using a CCD camera installed at the location of the tunnel that needs to be monitored to collect and store the original image; S2: reconstruction of the three-dimensional point cloud model of the original image, reconstruction of the three-dimensional model of the collected original image; S3: real-time image, using the CCD camera installed at the position of the tunnel to be monitored in S1 to take regular photos of the position of the tunnel and store them; S4: reconstruction of the three-dimensional point cloud model of the real-time image, reconstruction of the three-dimensional model of the collected real-time image; S5: Perform 3D model similarity topology comparison on the 3D point cloud reconstruction models in S2 and S4; S6: The difference between the early and late deformations of the tunnel is obtained through S5, so as to estimate the damage of the tunnel during use; In S1 and S3, the original image is collected and stored by using the light from the outside or the light reflected by the object to irradiate the CCD image sensor through the lens to obtain an electrical signal, and then the image signal processing circuit and the amplification circuit are used to generate a digital image of each row or multiple rows in coordination with the clock synchronization signal. The final image is stored on the memory card through the digital image, which is converted into a two-dimensional digital signal to generate a black and white image; a layer of red, green and blue film is covered on the sensor, and any color can be expressed through the three primary colors to simulate a color, so the color sensor generates color information; another layer of film is added on top, and a tiny lens is engraved by photolithography to ensure that the light can be focused on the small grid at the bottom through the lens, which can improve the sensor's sensitivity to light and conversion efficiency.

2. The tunnel surface deformation monitoring method based on multi-baseline close-range photogrammetry according to claim 1 is characterized in that: In S2 and S4, the reconstruction of the original image three-dimensional point cloud model includes the following steps: Step 1: Integrate the spatial point cloud data and color intensity data acquired by the RGB-D camera and manage and output them in an engineering manner; Step 2: Use PCL technology, a modular cross-platform open source C++ programming library for 3D point cloud processing, to pre-process the point cloud data, including denoising, segmentation, filtering, registration, and sampling operations, and output a point cloud with clear features and streamlined data; Step 3: Meshing the point cloud data, that is, using a series of grids to approximate the point cloud, using triangular grids and quadrilateral grids. This step realizes the conversion of point cloud to mesh in three-dimensional representation; Step 4: Panoramic texture mapping: Map the color and texture information collected by the RGB-D camera onto the mesh model, refine it, and output a realistic 3D model.

3. The tunnel surface deformation monitoring method based on multi-baseline close-range photogrammetry according to claim 1 is characterized in that: In S3, when the CCD camera installed at the position of the tunnel to be monitored takes timed photos of the position of the tunnel, the time of the timed photos can be set by people, specifically, 2 minutes, 5 minutes, 8 minutes and 10 minutes.

4. The tunnel surface deformation monitoring method based on multi-baseline close-range photogrammetry according to claim 1 is characterized in that: In S5, each three-dimensional value in the original image is defined as a target Oi...On, and each three-dimensional value of a feature point in the real-time image is defined as Di...Dn.

5. The tunnel surface deformation monitoring method based on multi-baseline close-range photogrammetry according to claim 4 is characterized in that: The three-dimensional coordinates (Xd, Yd, Nd) of the feature point of each real-time image are compared with the three-dimensional coordinates (Xi, Yi, Ni).

6. The tunnel surface deformation monitoring method based on multi-baseline close-range photogrammetry according to claim 5 is characterized in that: If the values ​​of the two are all equal, it means that the 3D point cloud models of the original image and the real-time image have not changed, so the deformation of the next frame of the tunnel can continue to be monitored.

7. The tunnel surface deformation monitoring method based on multi-baseline close-range photogrammetry according to claim 5 is characterized in that: If the two values ​​are not equal, Xd and Xi are equal, but Yd and Yi are not equal to Nd and Ni, that is, the two values ​​are not equal. Similarly, if two values ​​are equal and one value is not equal, it can also be defined as unequal, which means that the three-dimensional point cloud model of the original image and the real-time image has changed, so that the deformation of the tunnel can be judged. If the unequal values ​​are within the controllable range, the CCD camera can continue to monitor the deformation of the next frame of the tunnel.

8. The tunnel surface deformation monitoring method based on multi-baseline close-range photogrammetry according to claim 5 is characterized in that: If the values ​​of the two are not equal, Xd and Xi are equal, but Yd and Yi are not equal to Nd and Ni, that is, the values ​​of the two are not equal. Similarly, if two values ​​are equal and one value is not equal, it can also be defined as unequal, which means that the three-dimensional point cloud model of the original image and the real-time image has changed, so the deformation of the tunnel can be judged. If the unequal values ​​are not within the set range, the deformation of the tunnel cannot maintain the continued use of the tunnel, and the deformed position of the tunnel can be repaired and abandoned.

Images