Intelligent detection method and system for structural deformation and deviation based on machine vision

Abstract

The invention discloses an intelligent structural deformation and deviation detection method and system based on machine vision, and the method comprises the steps: obtaining the working data and state data of a detection target, carrying out the distortion processing of a camera image, obtaining a corrected image, processing the corrected image, obtaining the coordinate information of a target point, and carrying out the detection. Constructing a detection target three-dimensional model according to the target point coordinate information, comparing the model to obtain basic structure deformation and deviation, obtaining first structure deformation and deviation and second structure deformation and deviation according to the state data, and constructing a structure deformation and deviation intelligent detection model, and inputting working data and state data of a to-be-detected target into the structural deformation and deviation intelligent detection model to obtain a structural deformation and deviation detection result. The method not only can improve the precision of structural deformation and deviation detection, but also has good interpretability, and can be directly applied to a structural deformation and deviation intelligent detection system based on machine vision.

Description

Technical Field

[0001] The present invention relates to the field of nondestructive testing, and in particular to a machine vision-based intelligent detection method and system for structural deformation and deviation. Background Art

[0002] With the rapid development of buildings and infrastructure, the requirements for structural safety and stability are increasing. The detection of structural deformation and deviation is of great significance for preventing catastrophic accidents and protecting people's lives and property. At the same time, with the rapid development of machine vision and artificial intelligence technology, intelligent detection methods based on machine vision have gradually emerged, providing new solutions for the detection of structural deformation and deviation.

[0003] Traditional detection methods mostly rely on manual contact measurement and visual inspection, which have problems such as difficulty in application in large and high-rise buildings, time-consuming and labor-intensive measurement, and great influence of human factors. In addition, the influence of environmental status factors on the detection results is ignored, resulting in limited accuracy and reliability of the detection results. In addition, most of the existing detection methods based on machine vision only stay at the level of two-dimensional image processing, which is difficult to accurately reflect the true three-dimensional shape and deformation of the structure. By fully considering factors such as transmission influence, image conversion and environmental status data, combining machine vision technology, image recognition technology, three-dimensional reconstruction technology, BIM technology and machine learning technology, an accurate, efficient and fast machine vision-based intelligent detection method and system for structural deformation and deviation is designed to overcome the shortcomings of existing structural deformation and deviation detection methods and systems, realize efficient and high-precision structural deformation and deviation detection, and provide strong technical support for structural safety monitoring and evaluation in the fields of bridges, buildings, machinery, etc. Summary of the invention

[0004] The purpose of the present invention is to provide a method and system for intelligent detection of structural deformation and deviation based on machine vision.

[0005] To achieve the above object, the present invention is implemented according to the following technical solutions:

[0006] The present invention comprises the following steps:

[0007] Acquire working data and status data of the detection target, and pre-process the working data and the status data; the working data includes monitoring data and camera images; the status data includes environmental data and orientation data;

[0008] Performing distortion processing on the camera image to obtain a corrected image, performing image processing on the corrected image to obtain point cloud data, and obtaining target coordinate information according to the point cloud data and the corrected image;

[0009] Construct a three-dimensional model of the detection target according to the target coordinate information, establish a BIM model of the detection target, and compare the three-dimensional model with the BIM model to obtain basic structural deformation and deviation;

[0010] Determine a first structural deformation and deviation according to the environmental data and the basic structural deformation and deviation, and determine a second structural deformation and deviation according to the orientation data and the basic structural deformation and deviation;

[0011] A structural deformation and deviation intelligent detection model is constructed according to the structural deformation and deviation and the monitoring data, and the working data and state data of the target to be detected are input into the structural deformation and deviation intelligent detection model to obtain the structural deformation and deviation detection result.

[0012] Furthermore, the method of performing distortion processing on the camera image to obtain a corrected image includes:

[0013] The camera distortion model is used to correct the camera image coordinates. The expression is:

[0014] x corr =x(1+k1r 2 +k2r 4 +k3r 6 )+2p1xy+p2(r 2 +2x 2 )+δ-μ / x

[0015] y corr =y(1+k1r 2 +k2r 4 +k3r 6 )+p1(r 2 +2x 2 )+2p2xy+δ-μ / y

[0016] Where (x corr ,y corr ) is the position after correction, (x, y) is the position before correction, k1, k2, k3 are radial distortion correction parameters, p1, p2 are tangential distortion correction parameters, δ is the mean of directional correction, and μ is the edge distortion value;

[0017] Calibrate the camera parameters based on OpenCV functions to obtain the camera intrinsic parameter matrix, tangential distortion correction parameters, and radial distortion correction parameters;

[0018] Use a chessboard calibration plate for calibration: capture images of the calibration plate at multiple positions and angles, detect corner points to obtain coarse positioning, perform sub-pixel processing to obtain corner point coordinates, establish the relationship between chessboard points and image points to obtain intrinsic parameter calibration results and distortion correction parameters, and correct the image to obtain a corrected image.

[0019] Furthermore, the method for obtaining the target coordinate information includes:

[0020] Load the depth image of the rectified image, extract the depth image pixel value, and convert the depth image into point cloud data corresponding to the three-dimensional coordinates according to the depth camera-detection target distance, image pixel value and camera internal parameter matrix;

[0021] Construct a CNN machine learning model, input the corrected image and point cloud data into the CNN machine learning model, use the CNN machine learning model to perform image recognition and feature extraction on the corrected image to obtain the image coordinates of the target, use the CNN machine learning model to perform data preprocessing and data enhancement on the point cloud data, and match the point cloud data and image coordinates based on the CNN machine learning model to determine the coordinate information of the target.

[0022] Furthermore, a method for constructing a three-dimensional model of a detection target according to the target coordinate information includes:

[0023] Combined with the measuring instrument rotation point calculation algorithm and the dynamic local window algorithm, the coordinates of the target point on the attached wire are determined according to the coordinate information of the single observation point target point output by the CNN machine learning model;

[0024] Construct the spatial index structure of the point cloud data near each target point, including: calculating the normal vector of the point cloud data, calculating the vector space, inversely solving the Poisson equation through the gradient relationship, extracting the isosurface and reconstructing the point cloud model of the object being measured. The expression is:

[0025]

[0026] U=a1u 2 +a2uv+a3v 2 +a4u+a5v+a6

[0027] c g =(4a1a3-a2 2 ) / (a4 2 +a5 2 +1) 2

[0028]

[0029] in is the vector space of the target point q, t is a point in the vicinity of the target point, p is an arbitrary point, For different surface areas, is a smooth function, is the inward surface normal vector, U(u,v) is the quadratic surface equation, a1, a2, a3, a4, a5, a6 are the coefficients of the quadratic surface equation, c g is the Gaussian curvature, is the mean curvature;

[0030] The point cloud model of the object to be measured is meshed, including: constructing seed triangles and extracting boundary edges, determining adaptive neighborhood points according to the average distance density, calculating projection coordinates, and determining the optimal extension point according to the projection coordinates. The expression is:

[0031] Z:n x (XP x )+n y (YP y )+n z (ZP z )=0

[0032] Q t3 =[cosθ·I+(1-cosθ)kk T +sinθ·ROK]·Q t2

[0033] Where Z is the neighborhood Q t1 The point cloud coordinates of the midpoint of the inner boundary edge P(P x ,P y ,P z ) and the corresponding normal vector n P (n x ,n y ,n z ), Q t2 (X, Y, Z) is the tangent plane projection coordinate, Q t3 is the rotation coordinate after the projection coordinate is rotated, I is the 3D unit matrix, θ is the normal vector n P The angle with the normal vector of the XOY plane, k is the rotation axis, and is the normal vector n P Cross product with the normal vector of the XOY plane;

[0034] The hole repair of the point cloud model of the object to be tested is performed to obtain the three-dimensional model of the detection target, including: hole boundary recognition, determination of hole boundary direction, adding vertices, adding vertices according to normal vectors and curvature, adding vertex inspection and mesh smoothing. The expression is:

[0035]

[0036] 1 / σ+1 / ξ=ρ,σ>0,ξ<-ξ

[0037] Where N n is the normal vector of the newly added mesh vertex, N l is the normal vector of the adjacent triangle of the newly added mesh vertex, g is the number of adjacent triangles in a ring of vertices, λ is the shape factor, B is the centroid distance, P n To add mesh vertices, P n ″is the newly added mesh vertex after mesh smoothing, σ is the positive influencing factor of the shrinking process, ξ is the negative influencing factor of the stretching process, P u , P v is a ring neighborhood, P u ′ , P v ′ Add mesh vertex P n The neighborhood of a ring after contraction, h is the number of neighbors of a ring.

[0038] Furthermore, the method for obtaining the deformation and deviation of the basic structure includes:

[0039] Establish a BIM model of the detection target, and compare the three-dimensional model with the BIM model to obtain basic deformation and deviation

[0040] Import the building information of the inspection target into REVIT software to build the BIM model of the inspection target;

[0041] The 3D model and BIM model of the detection target are registered through grid nodes. By comparing the grid size and model reconstruction accuracy of each target point, the quantitative relationship of the deviation mode is determined. The expression is:

[0042] f(m,n,o)=f P (m,n,o)-f BIM (m,n,o)

[0043] Where f(m,n,o) is the detection target deviation field, m, n, o are the numbers of the target points in the x, y, and z directions, and f P (m,n,o) are the target point measurement values ​​on the 3D model, f BIM (m,n,o) are the theoretical values ​​of the target points on the BIM model;

[0044] Perform DCT forward transform on the detection target deviation field to obtain the orthogonal transform coefficient matrix, which is expressed as:

[0045]

[0046] Where C(μ,v,ω) is the orthogonal transformation coefficient matrix, M, N, O are the total number of sampling points in the x, y, z directions, and μ, v, ω are frequency values;

[0047] The key modes of the target deviation field are identified according to the orthogonal transformation coefficient matrix C(μ, v, ω), and important deviation data are selected according to the energy compaction and contribution of the deviation field. The expression is:

[0048]

[0049] Where E i is the energy compaction of the important deviation mode, Ci 2 (μ,v,ω) is the energy of the important deviation mode, C i (μ,v,ω)∈Ω, Ω is the modal set, η i is the contribution of the important deviation mode;

[0050] Select the deviation modes whose energy compaction and contribution are greater than the set threshold to form the key orthogonal transformation coefficient matrix Orthogonal transformation coefficient matrix Perform inverse DCT transform to obtain the key deviation field of the detection target, which is expressed as:

[0051]

[0052] in In order to detect the key deviation field of the target, the basic structure deformation and deviation are determined according to the key deviation field of the target.

[0053] Furthermore, the method for determining the first structural deformation and deviation includes:

[0054] The environmental data is input into the environmental impact function to obtain the environmental impact factor. The first deviation is determined according to the environmental impact factor and the basic deviation of the coordinates at each target point. The first structural deformation is determined according to the first deviation of the coordinates at the overall target point of the detection target. The expression is:

[0055]

[0056]

[0057] Effect1 is the environmental impact factor, w1, w2, w3, w4 are the weights of environmental impact factors, is the average light intensity, L Rm , L Gm , L Bm is the measured value of red, green and blue light in the environment, L Rs , L Gs , L Bs is the standard value of red, green and blue light in the environment, T m is the camera operating temperature, T s is the standard operating temperature of the camera, H m is the measured value of atmospheric humidity, H s is the standard value of atmospheric humidity, N m is the measured value of environmental noise, N m is the acceptable threshold of environmental noise, Deviation1 is the first deviation of the coordinates of the overall target point of the detection target, Deviation0 is the basic deviation of the coordinates of the overall target point of the detection target, and P n ″is the target point coordinate after grid smoothing, and P0 is the target point coordinate in the BIM model.

[0058] Furthermore, the method for determining the deformation and deviation of the second structure includes:

[0059] The scale influence factor is constructed according to the orientation data, and the second deviation is determined according to the scale influence factor and the basic deviation of the coordinates at each target point. The second structural deformation is determined according to the second deviation of the coordinates at the overall target point of the detection target. The expression is:

[0060]

[0061] Effect2 is the scale influence factor, w5, w6, w7 are the scale influence factor weights, D is the distance from the camera to the detection target, S BIM To detect the maximum values ​​of length, width, and height in the target BIM model, S0 is the standard size for camera detection. is the inclination angle between the camera and the object to be detected, φ is the rotation angle between the camera and the object to be detected, Deviation2 is the second deviation of the coordinates of the overall target point of the detection target, and w8 is the second structural deviation weight.

[0062] Furthermore, the method for obtaining the structural deformation and deviation detection result includes:

[0063] Measure the structural deformation and deviation of the detection target to obtain the actual structural deformation and deviation;

[0064] The actual structural deformation and deviation, the first structural deformation and deviation, the second structural deformation and deviation, the environmental influencing factor and the scale influencing factor are used to form a comprehensive detection data set, and the comprehensive detection data set is divided into a training set and a test set;

[0065] Construct a structural deformation and deviation intelligent detection model, which includes an attention layer, a feature fusion layer, and a BP neural network;

[0066] The attention layer extracts the main features of the training set data through two parallel SE attention mechanism channels, and learns the dependencies between the features to obtain the required feature data. The expression is:

[0067]

[0068] Among them, F D ′ eviation is the required feature data matrix, σ(·) is the sigmoid function used to compress the value between (0,1) to generate the weight coefficient, W1 and W2 are the weight matrices of the fully connected layer, which are used to compress and restore the number of channels C, δ(·) is the nonlinear activation function, I and J are the row and column dimensions of the feature data matrix, and fc (i, j) is the value of the cth channel of the feature data matrix at position (i, j), F Deviation It is the feature data matrix formed after the input data features are extracted;

[0069] The feature fusion layer performs probability calculation and feature fusion on the data processed by the attention layer through the first classification layer and the second classification layer, outputs the structural deformation and deviation fusion results of the detected target, uses the cross entropy loss function to optimize the classification results of the first classification layer, and uses the MSE function to optimize the fusion results of the second classification layer;

[0070] The BP neural network is used to learn the relationship between the structural deformation and deviation fusion results of the detection target and the actual structural deformation and deviation, environmental influencing factors, and scale influencing factors, and to perform regression prediction to output the structural deformation and deviation detection results of the detection target. The MAE loss function is used to uniformly punish the prediction error of the model, and the RMSProp optimizer is used to automatically adjust the learning rate of the model.

[0071] The test set is used to evaluate the intelligent detection model of structural deformation and deviation;

[0072] The working data and state data of the target to be detected are input into the structural deformation and deviation intelligent detection model to obtain the structural deformation and deviation detection results.

[0073] The second aspect is the intelligent detection system of structural deformation and deviation based on machine vision, including:

[0074] Data acquisition module: including a high-resolution binocular camera, a target and a measurement track; used to obtain the working data and status data of the detection target, and pre-process the working data and the status data;

[0075] Image module: used for performing distortion processing on the camera image to obtain a corrected image, performing image processing on the corrected image to obtain point cloud data, and obtaining target coordinate information according to the point cloud data and the corrected image;

[0076] BIM model module: used to establish the BIM model of the detection target based on the building information;

[0077] Data processing module: used for constructing a three-dimensional model of the detection target according to the target coordinate information, for comparing the three-dimensional model with the BIM model to obtain basic structural deformation and deviation, for constructing environmental impact factors and scale impact factors according to environmental data and orientation data, and for determining the first structural deformation and deviation and the second structural deformation and deviation according to the environmental impact factors, the scale impact factors and the basic structural deformation and deviation;

[0078] Detection model module: used to construct a structure deformation and deviation intelligent detection model according to the first structure deformation and deviation, the second structure deformation and deviation, the environmental influencing factor, the scale influencing factor and the actual structure deformation and deviation, perform feature fusion and regression prediction, and input the working data and state data of the target to be detected into the structure deformation and deviation intelligent detection model to obtain the structure deformation and deviation detection result;

[0079] Intelligent supervision module: used to store, view and manage the working data, the status data and the structural deformation and deviation detection results.

[0080] The beneficial effects of the present invention are:

[0081] The present invention is a method and system for intelligent detection of structural deformation and deviation based on machine vision. Compared with the prior art, the present invention has the following technical effects:

[0082] The present invention can improve the accuracy of structural deformation and deviation detection through image distortion processing and coordinate recognition, reconstruction of three-dimensional models, construction of environmental influencing factors and scale influencing factors, correction of structural deformation and deviation, and construction of model steps, thereby improving the efficiency and precision of structural deformation and deviation detection based on machine vision, and making structural deformation and deviation detection intelligent, which can greatly save resources and improve the efficiency of structural deformation and deviation detection, can realize the measurement of structural deformation and deviation of the target to be detected, and quickly perform high-precision structural deformation and deviation detection on the target to be detected, providing strong technical support for the construction field, and is of great significance for realizing ultra-high precision non-contact measurement of civil engineering structures, and can adapt to different structural deformation and deviation intelligent detection systems based on machine vision and different users' intelligent detection needs of structural deformation and deviation based on machine vision, and has certain universality. BRIEF DESCRIPTION OF THE DRAWINGS

[0083] Figure 1 The present invention is a flowchart of the steps of the intelligent detection method of structural deformation and deviation based on machine vision. DETAILED DESCRIPTION

[0084] The present invention is further described below by means of specific embodiments. The illustrative embodiments and descriptions of the present invention are used to explain the present invention but are not intended to limit the present invention.

[0085] The machine vision-based intelligent detection method and system for structural deformation and deviation of the present invention comprises the following steps:

[0086] like Figure 1 As shown, in this embodiment, the following steps are included:

[0087] Acquire working data and status data of the detection target, and pre-process the working data and the status data; the working data includes monitoring data and camera images; the status data includes environmental data and orientation data;

[0088] Performing distortion processing on the camera image to obtain a corrected image, performing image processing on the corrected image to obtain point cloud data, and obtaining target coordinate information according to the point cloud data and the corrected image;

[0089] Construct a three-dimensional model of the detection target according to the target coordinate information, establish a BIM model of the detection target, and compare the three-dimensional model with the BIM model to obtain basic structural deformation and deviation;

[0090] Determine a first structural deformation and deviation according to the environmental data and the basic structural deformation and deviation, and determine a second structural deformation and deviation according to the orientation data and the basic structural deformation and deviation;

[0091] A structural deformation and deviation intelligent detection model is constructed according to the structural deformation and deviation and the monitoring data, and the working data and state data of the target to be detected are input into the structural deformation and deviation intelligent detection model to obtain the structural deformation and deviation detection result.

[0092] In this embodiment, the method of performing distortion processing on the camera image to obtain a corrected image includes:

[0093] The camera distortion model is used to correct the camera image coordinates. The expression is:

[0094] x corr =x(1+k1r 2 +k2r 4 +k3r 6 )+2p1xy+p2(r 2 +2x 2 )+δ-μ / x

[0095] y corr =y(1+k1r 2 +k2r 4 +k3r 6 )+p1(r 2 +2x 2 )+2p2xy+δ-μ / y

[0096] Where (x corr ,y corr ) is the position after correction, (x, y) is the position before correction, k1, k2, k3 are radial distortion correction parameters, p1, p2 are tangential distortion correction parameters, δ is the mean of directional correction, and μ is the edge distortion value;

[0097] Calibrate the camera parameters based on OpenCV functions to obtain the camera intrinsic parameter matrix, tangential distortion correction parameters, and radial distortion correction parameters;

[0098] Calibration is performed using a chessboard calibration plate: images of the calibration plate are captured at multiple positions and angles, corner points are detected to obtain rough positioning, sub-pixel processing is performed to obtain corner point coordinates, the relationship between chessboard points and image points is established to obtain intrinsic parameter calibration results and distortion correction parameters, and the image is corrected to obtain a corrected image;

[0099] In the actual evaluation, the structural deformation and deviation of a gymnasium using a steel structure-concrete beam composite structure were detected, and the image coordinates of the key targets in the concrete beam image were obtained by a binocular camera: 1. (800, 600); 2. (1200, 400); 3. (500, 800); 4. (1500, 700); 5. (1000, 1000); 6. (1300, 500); 7. (700, 1100); 8. (1600, 300); 9. (900, 1200); 10. (1100, 600);

[0100] The intrinsic parameter matrix of the camera is [718.856, 0, 960.5; 0, 718.856, 540; 0, 0, 1], the distortion coefficient is [-0.2635, 0.05159, -0.0001, 0, 0], and the image coordinates of the key target after distortion processing are: 1. (802, 598); 2. (1203, 398); 3. (498, 802); 4. (1500, 700); 5. (1000, 1000); 6. (1300, 500); 7. (700, 1100); 8. (1600, 300); 9. (900, 1200); 10. (1100, 600).

[0101] In this embodiment, the method for obtaining the target coordinate information includes:

[0102] Load the depth image of the rectified image, extract the depth image pixel value, and convert the depth image into point cloud data corresponding to the three-dimensional coordinates according to the depth camera-detection target distance, image pixel value and camera internal parameter matrix;

[0103] Construct a CNN machine learning model, input the corrected image and point cloud data into the CNN machine learning model, use the CNN machine learning model to perform image recognition and feature extraction on the corrected image to obtain the image coordinates of the target, use the CNN machine learning model to perform data preprocessing and data enhancement on the point cloud data, and match the point cloud data and image coordinates based on the CNN machine learning model to determine the coordinate information of the target;

[0104] In the actual evaluation, image recognition is performed on the corrected image, and the coordinate information of the target is obtained by combining the distance between the camera and the detection target: top steel structure: 1. (205, -2, 10009); 2. (1503, 1, 10006); 3. (3498, 2, 10003); 4. (503, -1, 10002); 5. (4497, 1, 10007); lower concrete beam: 1. (2499, -1501, 7001); 2. (1501, -1499, 6999); 3. (3499, -1502, 7003); 4. (502, -1500, 6998); 5. (4499, -1501, 7002).

[0105] In this embodiment, the method for constructing a three-dimensional model of a detection target according to the target coordinate information includes:

[0106] Combining the measuring instrument rotation point calculation algorithm and the dynamic local window algorithm, the coordinates of the target point on the attached wire are determined according to the coordinate information of the single observation point target point output by the CNN machine learning model;

[0107] The kd-tree is used to construct the spatial index structure of the point cloud data near each target point, including: using the nearest neighbor search to calculate the normal vector of the point cloud data, calculating the vector space, inversely solving the Poisson equation through the gradient relationship, using the stereo cube algorithm to extract the isosurface and reconstruct the point cloud model of the object being measured. The expression is:

[0108]

[0109] U=a1u 2 +a2uv+a3v 2 +a4u+a5v+a6

[0110] c g =(4a1a3-a2 2 ) / (a4 2 +a5 2 +1) 2

[0111]

[0112] in is the vector space of the target point q, t is a point in the vicinity of the target point, p is an arbitrary point, For different surface areas, is a smooth function, is the inward surface normal vector, U(u,v) is the quadratic surface equation, a1, a2, a3, a4, a5, a6 are the coefficients of the quadratic surface equation, c g is the Gaussian curvature, is the mean curvature;

[0113] The point cloud model of the object to be measured is meshed, including: constructing seed triangles and extracting boundary edges, determining adaptive neighborhood points according to the average distance density, calculating projection coordinates, and determining the optimal extension point according to the projection coordinates. The expression is:

[0114] Z:n x (XP x )+n y (YP y )+n z (ZP z )=0

[0115] Q t3 =[cosθ·I+(1-cosθ)kk T +sinθ·ROK]·Q t2

[0116] Where Z is the neighborhood Q t1 The point cloud coordinates of the midpoint of the inner boundary edge P(P x ,P y ,P z ) and the corresponding normal vector n P (n x ,n y ,n z ), Q t2 (X, Y, Z) is the tangent plane projection coordinate, Q t3 is the rotation coordinate after the projection coordinate is rotated, I is the 3D unit matrix, θ is the normal vector n P The angle with the normal vector of the XOY plane, k is the rotation axis, and is the normal vector n P Cross product with the normal vector of the XOY plane;

[0117] The hole repair of the point cloud model of the object to be tested is performed to obtain the three-dimensional model of the detection target, including: hole boundary recognition, determination of hole boundary direction, adding vertices, adding vertices according to normal vectors and curvature, adding vertex inspection and mesh smoothing. The expression is:

[0118]

[0119]

[0120] 1 / σ+1 / ξ=ρ,σ>0,ξ<-ξ

[0121] Where N n is the normal vector of the newly added mesh vertex, N l is the normal vector of the adjacent triangle of the newly added mesh vertex, g is the number of adjacent triangles in a ring of vertices, λ is the shape factor, B is the centroid distance, P n To add mesh vertices, Pn ″ is the newly added mesh vertex after mesh smoothing, σ is the positive influencing factor of the shrinking process, ξ is the negative influencing factor of the stretching process, P u , P v is a ring neighborhood, P u ′ , P v ′ Add mesh vertex P n The neighborhood of a ring after contraction, h is the number of neighbors of a ring.

[0122] In this embodiment, the method for obtaining the basic structure deformation and deviation includes:

[0123] Establish a BIM model of the detection target, and compare the three-dimensional model with the BIM model to obtain basic deformation and deviation

[0124] Import the building information of the inspection target into REVIT software to build the BIM model of the inspection target;

[0125] The 3D model and BIM model of the detection target are registered through grid nodes. By comparing the grid size and model reconstruction accuracy of each target point, the quantitative relationship of the deviation mode is determined. The expression is:

[0126] f(m,n,o)=f P (m,n,o)-f BIM (m,n,o)

[0127] Where f(m,n,o) is the detection target deviation field, m, n, o are the numbers of the target points in the x, y, and z directions, and f P (m,n,o) are the target point measurement values ​​on the 3D model, f BIM (m,n,o) are the theoretical values ​​of the target points on the BIM model;

[0128] Perform DCT forward transform on the detection target deviation field to obtain the orthogonal transform coefficient matrix, which is expressed as:

[0129]

[0130] Where C(μ,v,ω) is the orthogonal transformation coefficient matrix, M, N, O are the total number of sampling points in the x, y, z directions, and μ, v, ω are frequency values;

[0131] The key modes of the target deviation field are identified according to the orthogonal transformation coefficient matrix C(μ, v, ω), and important deviation data are selected according to the energy compaction and contribution of the deviation field. The expression is:

[0132]

[0133] Where Ei is the energy compaction of the important deviation mode, C i 2 (μ,v,ω) is the energy of the important deviation mode, C i (μ,v,ω)∈Ω, Ω is the modal set, η i is the contribution of the important deviation mode;

[0134] Select the deviation modes whose energy compaction and contribution are greater than the set threshold to form the key orthogonal transformation coefficient matrix Orthogonal transformation coefficient matrix Perform inverse DCT transform to obtain the key deviation field of the detection target, which is expressed as:

[0135]

[0136] in To detect the target key deviation field, the basic structure deformation and deviation are determined according to the detected target key deviation field;

[0137] In the actual evaluation, the target coordinate information on the attached wire is determined by combining the binocular camera data of multiple measuring points. The three-dimensional information is reconstructed according to the target coordinate information to obtain the three-dimensional model of the detection target, and the BIM model is constructed at the same time. The original coordinates of the target points corresponding to the model are: top steel structure: 1. (2500, 0, 10000); 2. (1500, 0, 10000); 3. (3500, 0, 10000); 4. (500, 0, 10000); 5. (4500, 0, 10000); lower concrete beam: 1. (2500, -1500, 7000); 2. (1500, -1500, 7000); 3. (3500, -1500, 7000); 4. (500, -1500, 7000); 5. (4500, -1500, 7000);

[0138] The three-dimensional model of the detection target and the BIM model are compared to determine the coordinate deviation at the target point: top steel structure: 1. (5, -2, 9), 2. (3, 1, 6), 3. (-2, 2, 3), 4. (3, -1, 2), 5. (-3, 1, 7), lower concrete beam: 1. (-1, -1, 1), 2. (1, 1, -1), 3. (-1, -2, 2), 4. (2, 0, -2), 5. (-2, -1, 3), the target coordinate deviation reflects that the concrete beam has only slight deformation, and the deformation in each direction is within 2 mm. The steel structure support on the upper part of the concrete beam has small deformation, with slight deformation in the horizontal direction and the maximum deformation in the vertical direction of 9 mm. At the same time, the deformation of the steel structure in the middle of the span is larger than that at both ends.

[0139] In this embodiment, the method for determining the first structural deformation and deviation includes:

[0140] The environmental data is input into the environmental impact function to obtain the environmental impact factor. The first deviation is determined according to the environmental impact factor and the basic deviation of the coordinates at each target point. The first structural deformation is determined according to the first deviation of the coordinates at the overall target point of the detection target. The expression is:

[0141]

[0142] Effect1 is the environmental impact factor, w1, w2, w3, w4 are the weights of environmental impact factors, is the average light intensity, L Rm , L Gm , L Bm is the measured value of red, green and blue light in the environment, L Rs , L Gs , L Bs is the standard value of red, green and blue light in the environment, T m is the camera operating temperature, T s is the standard operating temperature of the camera, H m is the measured value of atmospheric humidity, H s is the standard value of atmospheric humidity, N m is the measured value of environmental noise, N m is the acceptable threshold of environmental noise, Deviation1 is the first deviation of the coordinates of the overall target point of the detection target, Deviation0 is the basic deviation of the coordinates of the overall target point of the detection target, and P n ″ is the target point coordinate after mesh smoothing, and P0 is the target point coordinate in the BIM model;

[0143] In the actual evaluation, the environmental data were obtained: the environmental RGB (245, 250, 253) when the camera was working, the environmental standard value RGB (255, 255, 255), the camera working temperature 31°C, the standard camera working temperature 25°C, the humidity in the gymnasium 60%, the standard atmospheric humidity 45%, the environmental noise measurement value 65dB, the camera working acceptable noise 45dB, w1 = 0.5, w2 = 0.6, w3 = 0.4, w4 = 0.5, and Effect1 = 0.177 was calculated based on the environmental data. The first structural deviation of the corresponding target point is: upper steel structure: 1.(7.02, -2.808, 12. 636), 2.(4.212,1.404,8.424), 3.(-2.808,2.808,4.212), 4.(4.212,-1.404,2.808), 5.(-4.212,1.404,9.828), lower concrete beam: 1.(-1.404,-1.404,1.404), 2.(1.404,1.404,-1.404), 3.(-1.404,-2.808,2.808), 4.(2.808,0,-2.808), 5.(-2.808,-1.404,4.212).

[0144] In this embodiment, the method for determining the deformation and deviation of the second structure includes:

[0145] The scale influence factor is constructed according to the orientation data, and the second deviation is determined according to the scale influence factor and the basic deviation of the coordinates at each target point. The second structural deformation is determined according to the second deviation of the coordinates at the overall target point of the detection target. The expression is:

[0146]

[0147] Effect2 is the scale influence factor, w5, w6, w7 are the scale influence factor weights, D is the distance from the camera to the detection target, S BIM To detect the maximum values ​​of length, width, and height in the target BIM model, S0 is the standard size for camera detection. is the inclination angle between the camera and the object to be detected, φ is the rotation angle between the camera and the object to be detected, Deviation2 is the second deviation of the coordinates of the overall target point of the detection target, and w8 is the second structural deviation weight;

[0148] In the actual evaluation, the orientation data were obtained: the shooting distance between the camera and the lower concrete beam was 5000 mm, the shooting angle was 30°, and the shooting angle was 5°; the shooting distance between the camera and the upper steel structure was 15000 mm, the shooting angle was 45°, and the shooting angle was -25°. w5 = 0.5, w6 = 0.5, w7 = 0.1, and w8 = 0.1. According to the environmental data, the Effect2 of the upper steel structure and the lower concrete were calculated to be 1.142 and 0.819 respectively. The second structural deviations of the corresponding targets were: upper steel structure: 1. (5.688, -1.763, 10.395), 2. (3 .365,1.087,6.858), 3.(-1.763,2.218,3.365), 4.(3.365,-0.924,2.218), 5.(-2.555,1.087,8.034), lower concrete beam: 1.(-0.940,-0.940,1.077), 2.(1.077,1.077,-0.940), 3.(-0.9404,-1.806,2.197), 4.(2.197,0,-1.806), 5.(-1.806,-0.940,3.333).

[0149] In this embodiment, the method for obtaining the structural deformation and deviation detection result includes:

[0150] Measure the structural deformation and deviation of the detection target to obtain the actual structural deformation and deviation;

[0151] The actual structural deformation and deviation, the first structural deformation and deviation, the second structural deformation and deviation, the environmental influencing factor and the scale influencing factor are used to form a comprehensive detection data set, and the comprehensive detection data set is divided into a training set and a test set;

[0152] Construct a structural deformation and deviation intelligent detection model, which includes an attention layer, a feature fusion layer, and a BP neural network;

[0153] The attention layer includes two parallel SE attention mechanism channels. The first structural deformation and deviation, the second structural deformation and deviation, and the corresponding environmental influence factors and scale influence factors in the training set are input into the two parallel SE attention mechanism channels respectively. The main features of the input data are extracted, and the dependency relationship between the features is learned to obtain the required feature data. The expression is:

[0154]

[0155] Among them, F D ′ eviationis the required feature data matrix, σ(·) is the sigmoid function used to compress the value between (0,1) to generate the weight coefficient, W1 and W2 are the weight matrices of the fully connected layer, which are used to compress and restore the number of channels C, δ(·) is the nonlinear activation function, I and J are the row and column dimensions of the feature data matrix, and f c (i, j) is the value of the cth channel of the feature data matrix at position (i, j), F Deviation It is the feature data matrix formed after the input data features are extracted;

[0156] The feature fusion layer includes the first classification layer and the second classification layer. The first classification layer is composed of concatenate fusion, batch normalization layer and softmax classifier. Concatenate concatenates data features along the axis to obtain representative features. The batch normalization layer normalizes the representative features. The softmax classifier converts the standardized representative features into probability results of each category. The second classification layer is composed of support vector machine SVM. The probability result is input for meta-classification to determine the fusion result of feature data. The cross entropy loss function is used to optimize the classification result of the first classification layer, and the MSE function is used to optimize the fusion result of the second classification layer.

[0157] The BP neural network is used to learn the relationship between the structural deformation and deviation fusion results of the detection target and the actual structural deformation and deviation, environmental influencing factors, and scale influencing factors, and to perform regression prediction to output the structural deformation and deviation detection results of the detection target. The MAE loss function is used to uniformly punish the prediction error of the model, and the RMSProp optimizer is used to automatically adjust the learning rate of the model.

[0158] The test set is used to evaluate the intelligent detection model of structural deformation and deviation;

[0159] Inputting the working data and status data of the target to be detected into the structural deformation and deviation intelligent detection model to obtain the structural deformation and deviation detection result;

[0160] In the actual evaluation, the working data and status data of the target to be detected are input into the structural deformation and deviation intelligent detection model to obtain the structural deformation and deviation detection results. Taking the detection results of the above-mentioned monitoring point camera as an example, the target deviations of the upper steel structure of the span are: 1. (5.903, -2.190, 10.677), 2. (3.526, 1.164, 7.091), 3. (-2.190, 2.342, 3.529), 4. (3.526, -1.109, 2.342), 5. (-3.256, 1.164, 8.287), the Target point deviations of the concrete beam at the lower part of the span: 1.(-1.115,-1.115,1.160), 2.(1.160,1.160,-1.115), 3.(-1.115,-2.205,2.335), 4.(2.335,0,-2.205), 5.(-2.205,-1.115,3.515). According to the target point coordinate deviation analysis, there is only slight deformation in the concrete beam at the lower part of the span, and there is small deformation in the steel structure above the concrete beam. The deformation at the middle of the span is greater than that at both ends, and the maximum deformation at the middle of the span reaches 10.677mm.

[0161] The second aspect is the intelligent detection system of structural deformation and deviation based on machine vision, including:

[0162] Data acquisition module: including a high-resolution binocular camera, a target and a measurement track; used to obtain the working data and status data of the detection target, and pre-process the working data and the status data;

[0163] Image module: used for performing distortion processing on the camera image to obtain a corrected image, performing image processing on the corrected image to obtain point cloud data, and obtaining target coordinate information according to the point cloud data and the corrected image;

[0164] BIM model module: used to establish the BIM model of the detection target based on the building information;

[0165] Data processing module: used for constructing a three-dimensional model of the detection target according to the target coordinate information, for comparing the three-dimensional model with the BIM model to obtain basic structural deformation and deviation, for constructing environmental impact factors and scale impact factors according to environmental data and orientation data, and for determining the first structural deformation and deviation and the second structural deformation and deviation according to the environmental impact factors, the scale impact factors and the basic structural deformation and deviation;

[0166] Detection model module: used to construct a structure deformation and deviation intelligent detection model according to the first structure deformation and deviation, the second structure deformation and deviation, the environmental influencing factor, the scale influencing factor and the actual structure deformation and deviation, perform feature fusion and regression prediction, and input the working data and state data of the target to be detected into the structure deformation and deviation intelligent detection model to obtain the structure deformation and deviation detection result;

[0167] Intelligent supervision module: used to store, view and manage the working data, the status data and the structural deformation and deviation detection results.

[0168] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. An intelligent detection method for structural deformation and deviation based on machine vision, characterized in that: The following steps are involved: S1, obtaining working data and status data of the detection target, and preprocessing the working data and the status data; the working data includes monitoring data and camera images; the status data includes environmental data and orientation data; S2, performing distortion processing on the camera image to obtain a corrected image, performing image processing on the corrected image to obtain point cloud data, and obtaining target coordinate information according to the point cloud data and the corrected image; S3, constructing a three-dimensional model of the detection target according to the target coordinate information, establishing a BIM model of the detection target, and comparing the three-dimensional model with the BIM model to obtain basic structural deformation and deviation; S4, determining a first structural deformation and deviation according to the environmental data and the basic structural deformation and deviation, and determining a second structural deformation and deviation according to the orientation data and the basic structural deformation and deviation; S5. Construct a structural deformation and deviation intelligent detection model according to the structural deformation and deviation and the monitoring data, and input the working data and state data of the target to be detected into the structural deformation and deviation intelligent detection model to obtain a structural deformation and deviation detection result.

2. According to claim 1, the method for intelligent detection of structural deformation and deviation based on machine vision is characterized in that: The method of performing distortion processing on the camera image to obtain a corrected image comprises: The camera distortion model is used to correct the camera image coordinates. The expression is: x corr =x(1+k1r 2 +k2r 4 +k3r 6 )+2p1xy+p2(r 2 +2x 2 )+δ-μ / x y corr =y(1+k1r 2 +k2r 4 +k3r 6 )+p1(r 2 +2x 2 )+2p2xy+δ-μ / y Where (x corr ,y corr ) is the position after correction, (x, y) is the position before correction, k1, k2, k3 are radial distortion correction parameters, p1, p2 are tangential distortion correction parameters, δ is the mean of directional correction, and μ is the edge distortion value; Calibrate the camera parameters based on OpenCV functions to obtain the camera intrinsic parameter matrix, tangential distortion correction parameters, and radial distortion correction parameters; Use a chessboard calibration plate for calibration: capture images of the calibration plate at multiple positions and angles, detect corner points to obtain coarse positioning, perform sub-pixel processing to obtain corner point coordinates, establish the relationship between chessboard points and image points to obtain intrinsic parameter calibration results and distortion correction parameters, and correct the image to obtain a corrected image.

3. According to claim 1, the method for intelligent detection of structural deformation and deviation based on machine vision is characterized in that: The method for obtaining the target coordinate information comprises: Load the depth image of the rectified image, extract the depth image pixel value, and convert the depth image into point cloud data corresponding to the three-dimensional coordinates according to the depth camera-detection target distance, image pixel value and camera internal parameter matrix; Construct a CNN machine learning model, input the corrected image and point cloud data into the CNN machine learning model, use the CNN machine learning model to perform image recognition and feature extraction on the corrected image to obtain the image coordinates of the target, use the CNN machine learning model to perform data preprocessing and data enhancement on the point cloud data, and match the point cloud data and image coordinates based on the CNN machine learning model to determine the coordinate information of the target.

4. According to claim 1, the method for intelligent detection of structural deformation and deviation based on machine vision is characterized in that: The method for constructing a three-dimensional model of a detection target according to the target point coordinate information comprises: Combining the measuring instrument rotation point calculation algorithm and the dynamic local window algorithm, the coordinates of the target point on the attached wire are determined according to the coordinate information of the single observation point target point output by the CNN machine learning model; Construct the spatial index structure of the point cloud data near each target point, including: calculating the normal vector of the point cloud data, calculating the vector space, inversely solving the Poisson equation through the gradient relationship, extracting the isosurface and reconstructing the point cloud model of the object being measured. The expression is: U=a1u 2 +a2uv+a3v 2 +a4u+a5v+a6 <h2 style=";text-align:left;direction:ltr">c<h2 style=";text-align:left;direction:ltr"> g <h2 style=";text-align:left;direction:ltr"> (4a1a3-a2)<h2 style=";text-align:left;direction:ltr"> 2 <h2 style=";text-align:left;direction:ltr"> ) / (a4<h2 style=";text-align:left;direction:ltr"> 2 <h2 style=";text-align:left;direction:ltr"> +a5<h2 style=";text-align:left;direction:ltr"> 2 <h2 style=";text-align:left;direction:ltr"> +1)<h2 style=";text-align:left;direction:ltr"> 2 in is the vector space of the target point q, t is a point in the vicinity of the target point, p is an arbitrary point, For different surface areas, is a smooth function, is the inward surface normal vector, U(u,v) is the quadratic surface equation, a1, a2, a3, a4, a5, a6 are the coefficients of the quadratic surface equation, c g is the Gaussian curvature, is the mean curvature; The point cloud model of the object to be measured is meshed, including: constructing seed triangles and extracting boundary edges, determining adaptive neighborhood points according to the average distance density, calculating projection coordinates, and determining the optimal extension point according to the projection coordinates. The expression is: Z:n x (XP x )+n y (YP y )+n z (ZP z )=0 Q t3 =[cosθ·I+(1-cosθ)kk T +sinθ·ROK]·Q t2 Where Z is the neighborhood Q t1 The point cloud coordinates of the midpoint of the inner boundary edge P(P x ,P y ,P z ) and the corresponding normal vector n P (n x ,n y ,n z ), Q t2 (X, Y, Z) is the tangent plane projection coordinate, Q t3 is the rotation coordinate after the projection coordinate is rotated, I is the 3D unit matrix, θ is the normal vector n P The angle with the normal vector of the XOY plane, k is the rotation axis, and is the normal vector n P Cross product with the normal vector of the XOY plane; The hole repair of the point cloud model of the object to be tested is performed to obtain the three-dimensional model of the detection target, including: hole boundary recognition, determination of hole boundary direction, adding vertices, adding vertices according to normal vectors and curvature, adding vertex inspection and mesh smoothing. The expression is: 1 / σ+1 / ξ=ρ,σ>0,ξ<-ξ Where N n is the normal vector of the newly added mesh vertex, N l is the normal vector of the adjacent triangle of the newly added mesh vertex, g is the number of adjacent triangles in a ring of vertices, λ is the shape factor, B is the centroid distance, P n To add mesh vertices, P n ″ is the newly added mesh vertex after mesh smoothing, σ is the positive influencing factor of the shrinking process, ξ is the negative influencing factor of the stretching process, P u , P v is a ring neighborhood, P u ′ , P v ′ Add mesh vertex P n The neighborhood of a ring after contraction, h is the number of neighbors of a ring.

5. According to claim 1, the method for intelligent detection of structural deformation and deviation based on machine vision is characterized in that: The method for obtaining the deformation and deviation of the basic structure comprises: Establish a BIM model of the detection target, and compare the three-dimensional model with the BIM model to obtain basic deformation and deviation Import the building information of the inspection target into REVIT software to build the BIM model of the inspection target; The 3D model and BIM model of the detection target are registered through grid nodes. By comparing the grid size and model reconstruction accuracy of each target point, the quantitative relationship of the deviation mode is determined. The expression is: f(m,n,o)=f P (m,n,o)-f BIM (m,n,o) Where f(m,n,o) is the detection target deviation field, m, n, o are the numbers of the target points in the x, y, and z directions, and f P (m,n,o) are the target point measurement values ​​on the 3D model, f BIM (m,n,o) are the theoretical values ​​of the target points on the BIM model; Perform DCT forward transform on the detection target deviation field to obtain the orthogonal transform coefficient matrix, which is expressed as: Where C(μ,v,ω) is the orthogonal transformation coefficient matrix, M, N, O are the total number of sampling points in the x, y, z directions, and μ, v, ω are frequency values; The key modes of the target deviation field are identified according to the orthogonal transformation coefficient matrix C(μ, v, ω), and important deviation data are selected according to the energy compaction and contribution of the deviation field. The expression is: Where E i is the energy compaction of the important deviation mode, C i 2 (μ,v,ω) is the energy of the important deviation mode, C i (μ,v,ω)∈Ω, Ω is the modal set, η i is the contribution of the important deviation mode; Select the deviation modes whose energy compaction and contribution are greater than the set threshold to form the key orthogonal transformation coefficient matrix Orthogonal transformation coefficient matrix Perform inverse DCT transform to obtain the key deviation field of the detection target, which is expressed as: in In order to detect the key deviation field of the target, the basic structure deformation and deviation are determined according to the key deviation field of the detected target.

6. The method for intelligent detection of structural deformation and deviation based on machine vision according to claim 1, characterized in that: The method for determining the first structural deformation and deviation comprises: The environmental data is input into the environmental impact function to obtain the environmental impact factor. The first deviation is determined according to the environmental impact factor and the basic deviation of the coordinates at each target point. The first structural deformation is determined according to the first deviation of the coordinates at the overall target point of the detection target. The expression is: Effect1 is the environmental impact factor, w1, w2, w3, w4 are the weights of environmental impact factors, is the average light intensity, L Rm , L Gm , L Bm is the measured value of red, green and blue light in the environment, L Rs , L Gs , L Bs is the standard value of red, green and blue light in the environment, T m is the camera operating temperature, T s is the standard operating temperature of the camera, H m is the measured value of atmospheric humidity, H s is the standard value of atmospheric humidity, N m is the measured value of environmental noise, N m is the acceptable threshold of environmental noise, Deviation1 is the first deviation of the coordinates of the overall target point of the detection target, Deviation0 is the basic deviation of the coordinates of the overall target point of the detection target, and P n ″ is the target point coordinate after grid smoothing, and P0 is the target point coordinate in the BIM model.

7. The method for intelligent detection of structural deformation and deviation based on machine vision according to claim 1, characterized in that: The method for determining the deformation and deviation of the second structure comprises: The scale influence factor is constructed according to the orientation data, and the second deviation is determined according to the scale influence factor and the basic deviation of the coordinates at each target point. The second structural deformation is determined according to the second deviation of the coordinates at the overall target point of the detection target. The expression is: Effect2 is the scale influence factor, w5, w6, w7 are the scale influence factor weights, D is the distance from the camera to the detection target, S BIM To detect the maximum values ​​of length, width, and height in the target BIM model, S0 is the standard size for camera detection, is the inclination angle between the camera and the object to be detected, φ is the rotation angle between the camera and the object to be detected, Deviation2 is the second deviation of the coordinates of the overall target point of the detection target, and w8 is the second structural deviation weight.

8. The method for intelligent detection of structural deformation and deviation based on machine vision according to claim 1, characterized in that: The method for obtaining the structural deformation and deviation detection result comprises: Measure the structural deformation and deviation of the detection target to obtain the actual structural deformation and deviation; The actual structural deformation and deviation, the first structural deformation and deviation, the second structural deformation and deviation, the environmental influencing factor and the scale influencing factor are used to form a comprehensive detection data set, and the comprehensive detection data set is divided into a training set and a test set; Construct a structural deformation and deviation intelligent detection model, which includes an attention layer, a feature fusion layer, and a BP neural network; The attention layer extracts the main features of the training set data through two parallel SE attention mechanism channels, and learns the dependencies between the features to obtain the required feature data. The expression is: where F D ′ eviation is the required feature data matrix, σ(·) is the sigmoid function used to compress the value between (0,1) to generate the weight coefficient, W1 and W2 are the weight matrices of the fully connected layer, which are used to compress and restore the number of channels C, δ(·) is the nonlinear activation function, I and J are the row and column dimensions of the feature data matrix, and f c (i, j) is the value of the cth channel of the feature data matrix at position (i, j), F Deviation It is the feature data matrix formed after the input data features are extracted; The feature fusion layer performs probability calculation and feature fusion on the data processed by the attention layer through the first classification layer and the second classification layer, outputs the structural deformation and deviation fusion results of the detected target, uses the cross entropy loss function to optimize the classification results of the first classification layer, and uses the MSE function to optimize the fusion results of the second classification layer; The BP neural network is used to learn the relationship between the structural deformation and deviation fusion results of the detection target and the actual structural deformation and deviation, environmental influencing factors, and scale influencing factors, and to perform regression prediction to output the structural deformation and deviation detection results of the detection target. The MAE loss function is used to uniformly punish the prediction error of the model, and the RMSProp optimizer is used to automatically adjust the learning rate of the model. The test set is used to evaluate the intelligent detection model of structural deformation and deviation; The working data and state data of the target to be detected are input into the structural deformation and deviation intelligent detection model to obtain the structural deformation and deviation detection results.

9. A machine vision-based intelligent detection system for structural deformation and deviation, used to execute the method according to any one of claims 1 to 8, characterized in that: include: Data acquisition module: including high-resolution binocular camera, target and measurement track; Used to obtain working data and status data of the detection target, and pre-process the working data and the status data; Image module: used for performing distortion processing on the camera image to obtain a corrected image, performing image processing on the corrected image to obtain point cloud data, and obtaining target coordinate information according to the point cloud data and the corrected image; BIM model module: used to establish the BIM model of the detection target based on the building information; Data processing module: used for constructing a three-dimensional model of the detection target according to the target coordinate information, for comparing the three-dimensional model with the BIM model to obtain basic structural deformation and deviation, for constructing environmental impact factors and scale impact factors according to environmental data and orientation data, and for determining the first structural deformation and deviation and the second structural deformation and deviation according to the environmental impact factors, the scale impact factors and the basic structural deformation and deviation; Detection model module: used to construct a structure deformation and deviation intelligent detection model according to the first structure deformation and deviation, the second structure deformation and deviation, the environmental influencing factor, the scale influencing factor and the actual structure deformation and deviation, perform feature fusion and regression prediction, and input the working data and state data of the target to be detected into the structure deformation and deviation intelligent detection model to obtain the structure deformation and deviation detection result; Intelligent supervision module: used to store, view and manage the working data, the status data and the structural deformation and deviation detection results.

Images