CN118111345A - Tunnel foundation pit surrounding rock displacement, crack and ponding monitoring system - Google Patents

Abstract

The present invention belongs to the technical field of engineering monitoring, and in particular relates to a monitoring system for displacement, cracks and water accumulation changes of surrounding rocks of a tunnel foundation pit based on laser radar and machine vision technology, comprising: a track laid on the top of a foundation pit; a mobile vehicle arranged on the track, which can move and inspect along the track; and a monitoring device, mounted on the mobile vehicle, for obtaining information about the side wall of the foundation pit inspected by the mobile vehicle; wherein the monitoring device uses a laser radar and a structured light camera to scan the side wall of the foundation pit to obtain three-dimensional point cloud data, and establishes a three-dimensional point cloud model of the foundation pit wall measurement in a laser radar coordinate system, and accurately measures the overall and local displacement deformation of the side wall of the foundation pit and the cracks and water seepage of the side wall through area division and data fitting.

Description

Tunnel foundation pit surrounding rock displacement, cracks and water accumulation monitoring system

Technical Field

The present invention belongs to the technical field of engineering monitoring, and in particular relates to a tunnel foundation pit surrounding rock displacement, crack and water accumulation change monitoring system based on laser radar and machine vision technology.

Background technique

With the development of society, issues such as environmental protection and safe production have received increasing attention. A large amount of electricity, energy, reinforced concrete and other resources will be consumed during the construction process, which will have a great impact on the environment. At the same time, the complexity and danger of the construction site determine the importance of safe construction. Therefore, the concept of green and safe construction is increasingly valued by the government, construction units, and construction units, and the corresponding national, local, and corporate policies and standards are constantly being introduced and updated. In this context, it is very necessary to carry out technical research and application for real-time safety monitoring and alarm of construction projects, and real-time monitoring and management of resources, environment, and safety information during the construction process.

As the difficulty of tunnel construction continues to increase, the stability detection of highway tunnel structures is particularly important. In traditional highway tunnel stability detection, visual inspection, destructive inspection or manual non-destructive testing methods are often used to detect tunnels. These methods are unstable and inefficient, so how to quickly and accurately detect tunnel stability is a research hotspot.

Summary of the invention

The purpose of the present invention is to propose a tunnel foundation pit surrounding rock displacement, cracks and water accumulation change monitoring system based on laser radar and machine vision technology to solve the above technical problems.

In order to achieve the above purpose, the following technical solution is provided: a tunnel foundation pit surrounding rock displacement, cracks and water accumulation change monitoring system, comprising:

Tracks, laid on top of the foundation pit;

A mobile trolley is arranged on a track and can perform mobile inspection along the track;

and a monitoring device, mounted on the mobile vehicle, for obtaining information about the side wall of the foundation pit inspected by the mobile vehicle;

Among them, the monitoring device uses laser radar and structured light camera to scan the side wall of the foundation pit to obtain three-dimensional point cloud data, and establishes a three-dimensional point cloud model of the foundation pit wall in the laser radar coordinate system. Through area division and data fitting, the overall and local displacement deformation of the side wall of the foundation pit, as well as the cracks and water seepage of the side wall, are accurately measured.

In the above technical solution, further, the monitoring device includes:

Stereo calibration module;

The image point cloud splicing and fusion module converts the image into 3D point cloud data, splices multiple frames of point cloud into a 3D model, smoothes and fuses the spliced point cloud, and finally processes the data to obtain the final 3D model.

The model fitting module selects the corresponding linear or nonlinear model according to the data characteristics and the target to be fitted; secondly, the least squares method, maximum likelihood estimation or other algorithms are used to calculate the parameter values of the model; the prediction performance of the model is evaluated using the validation data set, and the prediction error is calculated; if the prediction performance of the model does not meet the requirements, the model parameters are fine-tuned to improve its accuracy; the obtained fitting model is used to predict the value of the unknown data;

In the detection module, the data is first preprocessed, including noise removal and standardization; secondly, the artificial intelligence model is trained using a machine learning algorithm and its effectiveness is evaluated; then the model parameters are adjusted to improve its accuracy; the model is tested on an independent data set and its accuracy is evaluated.

In any of the above technical solutions, further, the stereo calibration module includes:

An image acquisition unit uses a structured light binocular camera to acquire images of the calibration surface;

A camera correction unit, which corrects the acquired image;

The point cloud construction unit uses the disparity map and the depth map to perform three-dimensional reconstruction of the feature points on the calibration surface to construct a point cloud;

The parameter calibration unit analyzes the point cloud data and calculates the camera's intrinsic and extrinsic parameters;

The camera calibration unit calibrates the internal and external parameters of the camera through the intrinsic parameters and extrinsic parameters calculated by the parameter calibration unit. The intrinsic parameters are the camera matrix and distortion coefficients, and the extrinsic parameters are the rotation matrix and translation vector.

In any of the above technical solutions, further, the data evaluation unit evaluates the calibration result to ensure the accuracy of the calibration.

In any of the above technical solutions, further, in the process of monitoring the displacement change of the side wall of the foundation pit, the side wall data of the foundation pit is obtained by multiple measurements through a laser radar, and then the noise data is filtered according to the initial parameters of the foundation pit measurement, and the side wall data is interpolated accordingly; the position change of the side wall is monitored by comparing the shooting height with the change of the side wall data under multiple measurements at a uniform height, and the displacement of the side wall is calculated.

In any of the above technical solutions, further, in the process of monitoring the cracks on the side wall of the foundation pit, a camera is used to collect data on the cracks on the side wall of the foundation pit; the original image is first cut into k×k small blocks by image cutting, and then the model training is performed on the image size;

Among them, the crack height positioning algorithm adopts the SVR regression algorithm. By constructing a fitting relationship function between the parameters of the foundation pit side wall, the height taken by the camera, the coordinates of the pixel points in the picture and the actual height corresponding to the pixel points in the picture, the heights corresponding to the four corner points of the different crack detection frames in the test picture are calculated, and then the height range of the crack is calculated.

In any of the above technical solutions, further, the monitoring device uses the RGB image and depth image of the fixed point reference target at the top of the foundation pit taken multiple times by the laser radar and the structured light camera to extract multiple groups of three-dimensional feature points in the image, calculates the pose transformation matrix of the structured light camera under multiple measurements relative to the first shooting, and combines the joint calibration algorithm of the structured light camera and the laser radar to unify the laser radar coordinate system under multiple measurements to the laser radar coordinate system under the first scan, thereby realizing the unification of the coordinate systems of multiple measurements.

In any of the above technical solutions, further, the joint calibration algorithm adopts a plane constraint-based calibration method to jointly calibrate the lidar and structured light camera.

In any of the above technical solutions, further, the camera calculates the coordinates of the calibration plate plane in the camera coordinate system by identifying the label on the calibration plate; the light beam emitted by the laser radar hits the calibration plate, and the plane constraints are constructed by using the coordinates of the laser point in the laser coordinate system and the coordinates in the camera coordinate system to solve the external parameters.

The plane constraint calculation formula is as follows:

n ^cT (RP ^l +t)+d ^c =0

Among them, vector n is the three-dimensional normal vector of the label plane, d is the distance from the origin of the camera coordinate system O _c to the label plane, the coordinates of point P in the laser coordinate system O _l are P ^l , and the coordinates in the camera coordinate system O _c are P ^c ; point P is transformed from the laser coordinate system O _l to the camera coordinate system O _c :

P ^c = ^RT (P ^l -t)

Substituting into P ^l =(x, y, 0) ^T :

The matrix H is treated as a new unknown quantity and a nonlinear least squares solution is performed to recover and obtain the rotation matrix and translation vector R_T′.

The beneficial effects of the present invention are as follows: the method adopts stereo calibration technology, image stitching technology, point cloud stitching technology, image and point cloud fusion technology, linear and nonlinear fitting technology, artificial intelligence detection technology, etc., to perform three-dimensional modeling calculation on the side wall of the foundation pit in different time periods, calculates the displacement and deformation of the side wall according to the position changes of feature points in different time periods, and adopts artificial intelligence detection technology to identify and detect cracks and water seepage in the side wall, so as to realize real-time monitoring of the side wall of the foundation pit.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG1 is a flow chart of constructing a point cloud of a target area according to the present invention;

FIG2 is a schematic diagram of the joint calibration of a camera and a laser radar according to the present invention;

Detailed ways

The following will be combined with the drawings in the embodiments of the present application to clearly describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments in the present application belong to the scope of protection of this application.

In the description of the present application, it should be noted that the terms used herein are only for describing specific embodiments, and are not intended to limit the exemplary embodiments according to the present application. For ease of description, the sizes of the various parts shown in the drawings are not drawn according to the actual proportional relationship. The technology, methods and equipment known to ordinary technicians in the relevant field may not be discussed in detail, but in appropriate cases, the technology, methods and equipment should be regarded as part of the authorization specification. In all examples shown and discussed here, any specific value should be interpreted as merely exemplary, rather than as a limitation. Therefore, other examples of the exemplary embodiments may have different values. It should be noted that similar numbers and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further discussed in subsequent drawings.

The displacement intelligent visual monitoring, analysis and early warning system belongs to the field of engineering monitoring. It mainly includes adjustable targets, intelligent visual recognition modules, intelligent acquisition and control modules, cloud platforms, clients and other parts.

Among them, the adjustable target is fixed on the top of the foundation pit and arranged opposite to the intelligent visual recognition module; the intelligent visual recognition module is connected to the intelligent acquisition control module through a wire; the intelligent acquisition control module is connected to the cloud platform through a network; the cloud platform parses and calculates the received data, compares it with the set allowable deformation threshold, and marks different threshold intervals with corresponding colors, using the different colors of the measuring points to represent the safety status of the measuring points, and sends alarm information to the designated recipient; finally, the deformation and warning information of each measuring point are intuitively visualized using the client.

The existing method is based on a single visual monitoring analysis, which has high requirements for the on-site environment and is easily affected by dust caused by on-site construction and personnel activities. In addition, the scheme must use targets as the measured points, which has high requirements for point selection and installation, and is greatly affected by construction. For example, construction machinery and vehicles block the targets, which will lead to a decrease in monitoring effect, and the monitoring items are also relatively single. Based on this, the applicant now provides a tunnel foundation pit surrounding rock displacement, cracks and water accumulation change monitoring system based on laser radar and machine vision technology.

The monitoring system of the present application is described in detail through the following embodiments.

Embodiment 1:

This embodiment provides a monitoring system for displacement, cracks and water accumulation of surrounding rock in a tunnel foundation pit, including: a track laid on the top of the foundation pit; a mobile vehicle arranged on the track, which can move and inspect along the track; and a monitoring device mounted on the mobile vehicle, used to obtain information about the side wall of the foundation pit inspected by the mobile vehicle;

Among them, the monitoring device uses lidar and structured light camera to scan the side walls of the foundation pit to obtain three-dimensional point cloud data, and establishes a three-dimensional point cloud model of the foundation pit wall in the lidar coordinate system. Through area division and data fitting, the overall and local displacement deformation of the side walls of the foundation pit, as well as cracks and water seepage in the side walls, are accurately measured.

Specifically, the monitoring device includes:

Stereo calibration module; The stereo calibration module includes: an image acquisition unit, which uses a structured light binocular camera to acquire images of the calibration surface; a camera correction unit, which corrects the acquired images; a point cloud construction unit, which uses a disparity map and a depth map to perform three-dimensional reconstruction of the feature points on the calibration surface to construct a point cloud; a parameter calibration unit, which analyzes the point cloud data and calculates the camera's intrinsic and extrinsic parameters; the camera calibration unit, which uses the intrinsic and extrinsic parameters calculated by the parameter calibration unit to calibrate the camera's internal and external parameters, the intrinsic parameters are the camera matrix and the distortion coefficient, and the extrinsic parameters are the rotation matrix and the translation vector; and a data evaluation unit, which evaluates the calibration results to ensure the accuracy of the calibration; The purpose of point cloud surface reconstruction is to enable it to vividly reflect the geometric features of the object surface, so as to facilitate the extraction and analysis of the model's physical and geometric information;

The image point cloud splicing and fusion module converts the image into 3D point cloud data, splices multiple frames of point cloud into a 3D model, smoothes and fuses the spliced point cloud, and finally processes the data to obtain the final 3D model.

The model fitting module selects the corresponding linear or nonlinear model according to the data characteristics and the target to be fitted; secondly, the least squares method, maximum likelihood estimation or other algorithms are used to calculate the parameter values of the model; the prediction performance of the model is evaluated using the validation data set, and the prediction error is calculated; if the prediction performance of the model does not meet the requirements, the model parameters are fine-tuned to improve its accuracy; the obtained fitting model is used to predict the value of the unknown data;

In the detection module, the data is first preprocessed, including noise removal and standardization; secondly, the artificial intelligence model is trained using a machine learning algorithm and its effectiveness is evaluated; then the model parameters are adjusted to improve its accuracy; the model is tested on an independent data set and its accuracy is evaluated.

In the present technical solution, the working principle of the laser radar for collecting point cloud information is: first, the laser transmitter emits a high-energy laser beam to the observed target entity through the observation lens. The laser is reflected after reaching the surface of the object, returns along the same path, and is received by the scanner receiver, thereby obtaining the distance between the scanner and the object being measured; then, the angle measurement control module is used to obtain the horizontal and vertical angle information between the target object and the scanner, and the drive motor and steerable lens are used to complete the collection of target point information; finally, the microprocessor is used to convert the collected distance information and angle information into coordinate information, and the coordinate information, reflection intensity, texture features and other information are stored in the memory.

Due to the influence of the laser radar itself, the external environment, human operation and other factors, there will be noise points in the collected point cloud data of the foundation pit side wall that interfere with the geometric feature information of the object. The side wall point cloud data that has not been denoised will not only affect the accuracy of subsequent point cloud data registration and the establishment of the side wall three-dimensional solid model, but also affect the subsequent dynamic deformation monitoring of the side wall. For this reason, denoising is required.

Large-scale noise in point cloud data mainly refers to those useless information points that are far away from the main point cloud and easy to distinguish, including isolated points distributed near the main point cloud and drift points that are far away from the main point cloud data and have a small number.

There are also small-scale noise points in the point cloud data collected by LiDAR caused by the external environment, which are mixed with valid data points or are close to the main point cloud data and difficult to distinguish. The model surface built with point cloud data containing small-scale noise is not smooth, rough and has poor effect, so it is necessary to remove or filter these points mixed in the point cloud data.

To this end, in the process of monitoring the displacement changes of the side walls of the foundation pit, the side wall data of the foundation pit are obtained by multiple measurements through lidar, and then the noise data is filtered according to the initial parameters of the foundation pit measurement (such as the initial distance between the instrument and the side wall, the height of the foundation pit, etc.), and the side wall data are interpolated accordingly to make the side wall data denser; the position changes of the side wall are monitored according to the changes in the side wall data under multiple measurements at the same height compared with the shooting height, and the displacement of the side wall is calculated.

In the process of monitoring the cracks on the side walls of foundation pits, the size of the original images collected is too large to be directly used as the input of the network model. Therefore, a camera is used to collect the data of the cracks on the side walls of foundation pits. The original images are first cut into k×k small blocks using image cutting, and then the model is trained on the image size.

Since the foundation pit crack detection model only obtains the position of the crack in the plane image, it cannot obtain the specific position of the crack on the side wall of the foundation pit, which is not conducive to the subsequent repair work. Therefore, after identifying the crack in the plane image, it is necessary to clarify the height of the identified crack. The SVR regression algorithm is used for the crack height positioning algorithm. By constructing the fitting relationship function of the foundation pit side wall parameters (height), the height captured by the camera, the coordinates of the pixel points in the picture and the actual height corresponding to the pixel points in the picture, the heights corresponding to the four corner points of the different crack detection boxes in the test picture are obtained, and then the height range of the crack is obtained.

In monitoring the water seepage in cracks on the side walls of foundation pits, the same image recognition is used, and the water seepage position and height are located through the water seepage recognition algorithm.

In this embodiment, the optimized monitoring device uses the RGB images and depth images of the fixed point reference targets on the top of the foundation pit taken multiple times by the laser radar and the structured light camera to extract multiple groups of three-dimensional feature points in the image, calculates the pose transformation matrix of the structured light camera under multiple measurements relative to the first shooting, and combines the joint calibration algorithm of the structured light camera and the laser radar to unify the laser radar coordinate system under multiple measurements to the laser radar coordinate system under the first scan, thereby realizing the unification of the coordinate systems of multiple measurements.

Since the RGB image is a projection of a three-dimensional scene on a two-dimensional plane and does not record depth information, in order to determine the position and posture of the target in three-dimensional space, it is necessary to align the depth image to the RGB image, construct a point cloud image, and finally determine the spatial position and posture of the target based on the point cloud processing related algorithm. As shown in Figure 1, a point cloud construction flow chart of the target area is constructed.

The method for extracting image feature points and calculating the rotation matrix is as follows:

SIFT feature is a local feature extraction method for images, which has scale invariance, rotation invariance and certain illumination invariance. The specific steps of the algorithm are as follows.

1) Scale space extreme value detection

Create a Gaussian pyramid of the image and calculate the Gaussian difference image (DoG) at different scales. Detect local extreme points in the DoG image, which may be potential key points. The Gaussian kernel function is shown as follows:

L(x, y, σ) = G(x, y, σ) * I(x, y)

Where G(x, y, σ) is the Gaussian kernel function:

2) Key point positioning

In the DoG algorithm, a poor key point has a larger principal curvature in the direction parallel to the edge and a smaller curvature in the direction perpendicular to the edge. If the ratio of the two is higher than a certain threshold, the key point is considered to be a boundary and will be ignored.

3) Determine the direction of key points

Each key point is assigned a direction to achieve rotation invariance. For any key point, the gradient features (amplitude and angle) of all pixels in the area with radius r in the Gaussian pyramid image are collected.

4) Key point description

The image area around the key point is divided into blocks, the gradient histogram within the block is calculated, and a feature vector is generated to abstract the image information.

According to the image processing, multiple three-dimensional feature point pairs of the target area at different time points are extracted: P = {p ₁ , ..., p _n }P′ = {p′ ₁ , ..., p′ _n }, and through the Euclidean transformation R_T, p _i = Rp _i +T. This problem can be solved by iterative closest point (ICP). Specifically, linear algebra solution (SVD) is used. By defining the error term, the least squares problem is constructed to find R_T that minimizes the sum of squared errors. (ICP) can be solved in the following three steps:

Step 1: Calculate the centroid positions p and p′ of the two groups of points, and then calculate the centroid coordinates of each point q _i = p _i -p, q′ _i = p′ _i -p′;

Step 2: Define the optimization problem and calculate the rotation matrix

Step 3: Calculate the translation variable

t=p-Rp′.

Regarding the joint calibration of structured light cameras and lidar, specifically, the joint calibration algorithm uses a plane constraint-based calibration method to jointly calibrate lidar and structured light cameras.

As shown in Figure 2, the camera calculates the coordinates of the calibration plate plane in the camera coordinate system by identifying the label on the calibration plate; the light beam emitted by the laser radar hits the calibration plate, and the plane constraints are constructed using the coordinates of the laser point in the laser coordinate system and the coordinates in the camera coordinate system to solve the external parameters.

The plane constraint calculation formula is as follows:

n ^cT (RP ^l +t)+d ^c =0

Among them, vector n is the three-dimensional normal vector of the label plane, d is the distance from the origin of the camera coordinate system O _c to the label plane, the coordinates of point P in the laser coordinate system O _l are P ^l , and the coordinates in the camera coordinate system O _c are P ^c ; point P is transformed from the laser coordinate system O _l to the camera coordinate system O _c :

P ^c = ^RT (P ^l -t)

Substituting into P ^l =(x, y, 0) ^T :

The matrix H is treated as a new unknown quantity and a nonlinear least squares solution is performed to recover and obtain the rotation matrix and translation vector R_T′.

The embodiments of the present application are described above in conjunction with the accompanying drawings. In the absence of conflict, the embodiments in the present application and the features in the embodiments can be combined with each other. The present application is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present application, ordinary technicians in this field can also make many forms without departing from the purpose of the present application and the scope of protection of the claims, all of which are within the protection of the present application.

Claims (9)

1. A tunnel foundation pit surrounding rock displacement, cracks and water accumulation change monitoring system, characterized by comprising: Tracks, laid on top of the foundation pit; A mobile trolley is arranged on a track and can perform mobile inspection along the track; and a monitoring device, mounted on the mobile vehicle, for obtaining information about the side wall of the foundation pit inspected by the mobile vehicle; Among them, the monitoring device uses laser radar and structured light camera to scan the side wall of the foundation pit to obtain three-dimensional point cloud data, and establishes a three-dimensional point cloud model of the foundation pit wall in the laser radar coordinate system. Through area division and data fitting, the overall and local displacement deformation of the side wall of the foundation pit, as well as the cracks and water seepage of the side wall, are accurately measured. 2. A tunnel foundation pit surrounding rock displacement, cracks and water accumulation change monitoring system according to claim 1, characterized in that the monitoring device comprises: Stereo calibration module; The image point cloud splicing and fusion module converts the image into 3D point cloud data, splices multiple frames of point cloud into a 3D model, smoothes and fuses the spliced point cloud, and finally processes the data to obtain the final 3D model. The model fitting module selects the corresponding linear or nonlinear model according to the data characteristics and the target to be fitted; secondly, the least squares method, maximum likelihood estimation or other algorithms are used to calculate the parameter values of the model; the prediction performance of the model is evaluated using the validation data set, and the prediction error is calculated; if the prediction performance of the model does not meet the requirements, the model parameters are fine-tuned to improve its accuracy; the obtained fitting model is used to predict the value of the unknown data; In the detection module, the data is first preprocessed, including noise removal and standardization; secondly, the artificial intelligence model is trained using a machine learning algorithm and its effectiveness is evaluated; then the model parameters are adjusted to improve its accuracy; the model is tested on an independent data set and its accuracy is evaluated. 3. According to the machine vision technology-based foundation pit surrounding rock displacement, crack and water accumulation monitoring system of claim 2, it is characterized in that the three-dimensional calibration module comprises: An image acquisition unit uses a structured light binocular camera to acquire images of the calibration surface; A camera correction unit, which corrects the acquired image; The point cloud construction unit uses the disparity map and the depth map to perform three-dimensional reconstruction of the feature points on the calibration surface to construct a point cloud; The parameter calibration unit analyzes the point cloud data and calculates the camera's intrinsic and extrinsic parameters; The camera calibration unit calibrates the internal and external parameters of the camera through the intrinsic parameters and extrinsic parameters calculated by the parameter calibration unit. The intrinsic parameters are the camera matrix and distortion coefficients, and the extrinsic parameters are the rotation matrix and translation vector. 4. According to claim 3, a system for monitoring changes in surrounding rock displacement, cracks and water accumulation in a foundation pit based on machine vision technology is characterized in that it also includes: a data evaluation unit that evaluates the calibration results to ensure the accuracy of the calibration. 5. A tunnel foundation pit surrounding rock displacement, cracks and water accumulation change monitoring system according to claim 4, characterized in that: In the process of monitoring the displacement changes of the foundation pit side wall, the laser radar is used to obtain the foundation pit side wall data under multiple measurements, and then the noise data is filtered according to the initial parameters of the foundation pit measurement, and the side wall data is interpolated accordingly; the position change of the side wall is monitored by comparing the changes in the side wall data under multiple measurements at the same height with the shooting height, and the side wall displacement is calculated. 6. A tunnel foundation pit surrounding rock displacement, cracks and water accumulation change monitoring system according to claim 4, characterized in that: In the process of monitoring the cracks on the side wall of the foundation pit, a camera is used to collect the crack data on the side wall of the foundation pit; the original image is first cut into k×k small blocks by image cutting, and then the model training is performed on the image size; Among them, the crack height positioning algorithm adopts the SVR regression algorithm. By constructing a fitting relationship function between the parameters of the foundation pit side wall, the height taken by the camera, the coordinates of the pixel points in the picture and the actual height corresponding to the pixel points in the picture, the heights corresponding to the four corner points of the different crack detection frames in the test picture are calculated, and then the height range of the crack is calculated. 7. A tunnel foundation pit surrounding rock displacement, cracks and water accumulation change monitoring system according to claim 5, characterized in that: The monitoring device uses the RGB images and depth images of the fixed point reference targets at the top of the foundation pit taken multiple times by the laser radar and structured light camera to extract multiple groups of three-dimensional feature points in the image, calculates the posture transformation matrix of the structured light camera under multiple measurements relative to the first shooting, and combines the joint calibration algorithm of the structured light camera and the laser radar to unify the laser radar coordinate system under multiple measurements to the laser radar coordinate system under the first scan, thereby realizing the unification of the coordinate systems of multiple measurements. 8. A tunnel foundation pit surrounding rock displacement, cracks and water accumulation change monitoring system according to claim 7, characterized in that: The joint calibration algorithm uses a plane constraint-based calibration method to jointly calibrate the laser radar and structured light camera. 9. A tunnel foundation pit surrounding rock displacement, cracks and water accumulation change monitoring system according to claim 8, characterized in that: The camera calculates the coordinates of the calibration plate plane in the camera coordinate system by identifying the label on the calibration plate; the light beam emitted by the laser radar hits the calibration plate, and the plane constraints are constructed using the coordinates of the laser point in the laser coordinate system and the coordinates in the camera coordinate system to solve the external parameters. The plane constraint calculation formula is as follows: n ^cT (RP ^l +t)+d ^c =0 Among them, vector n is the three-dimensional normal vector of the label plane, d is the distance from the origin of the camera coordinate system O _c to the label plane, the coordinates of point P in the laser coordinate system O _l are P ^l , and the coordinates in the camera coordinate system O _c are P ^c ; point P is transformed from the laser coordinate system O _l to the camera coordinate system O _c : P ^c = ^RT (P ^l -t) Substituting into P ^l =(x, y, 0) ^T : The matrix H is treated as a new unknown quantity and a nonlinear least squares solution is performed to recover and obtain the rotation matrix and translation vector R_T′.