A method and system for periodic monitoring of parameters of developing cracks

Abstract

The present invention discloses a parameter periodic monitoring method and system for developing cracks. The method comprises the following steps: first, the histogram shape of the crack images taken at different times is adjusted to be consistent by using a histogram normalization method to improve the consistency and reliability of the detection results; then, in order to solve the spatial resolution difference caused by the inconsistent camera posture at different times, a calibration plate is placed in the crack area, and the geometric features and scale information of the crack are restored by affine correction and metric correction methods; finally, in order to determine the precise position of the specified section in the sequence image, a digital image correlation method is used to match and locate the crack area in the sequence image. The present invention solves the technical difficulties of insufficient measurement accuracy caused by illumination changes, spatial resolution differences and crack section positioning errors in the periodic monitoring of crack parameters, and can provide accurate data support for the growth and evolution of cracks and the health assessment of engineering structures.

Description

Technical Field

[0001] The invention relates to a method for monitoring crack parameters, and in particular to a method and system for periodic monitoring parameters of developing cracks. Background Art

[0002] Cracks can significantly reduce the integrity and bearing capacity of engineering structures. Without obvious environmental or human disturbances, crack expansion caused by material aging, stress changes, and temperature and humidity fluctuations is one of the main threats to the safety of engineering structures. Crack parameters can accurately reflect the degree of damage to the engineering structure, where the width reflects the degree of damage to the structure, and the length and direction reflect the distribution pattern of the structural load. Accurate measurement of crack parameters can provide a scientific basis for maintenance plans, reasonably arrange maintenance time and resources, avoid sudden failure of the structure during use, and ensure the long-term safe operation of the structure.

[0003] For mature cracks, only one test is needed to determine the crack parameters and assess their impact on the health of the structure. In contrast, developing cracks often require multiple and continuous tests. This is because such cracks are in a state of change, and their parameters such as shape, length, and width are not fixed. The data obtained from a single test has limitations and it is difficult to fully reflect their development trends and potential hazards. Ideally, the parameters of developing cracks should be monitored for a long time, but given that crack development is usually slow, and considering factors such as equipment stability, work efficiency, and waste of computing resources, the application of long-term monitoring in actual engineering is also subject to many constraints. Periodic monitoring based on computer vision can not only detect the development trend of cracks in a timely manner, but also realize non-destructive and high-precision calculation of crack parameters. Therefore, the periodic monitoring method is one of the best strategies for the state assessment and evolution trend exploration of developing cracks. Through multiple tests and comprehensive analysis of the series of data obtained, the evolution trend of cracks can be accurately grasped, so as to more effectively assess their threats to the health of the structure and provide strong support for taking timely and appropriate repair and protection measures.

[0004] The crack parameter monitoring method based on computer vision is a method that uses digital image processing and pattern recognition methods to automatically analyze crack images to achieve crack location and parameter quantification. This method mainly uses image processing methods (such as denoising, edge detection, threshold segmentation, skeleton extraction, etc.) to automatically identify and analyze crack features through the surface image of the engineering structure. Compared with traditional methods, this method has significant advantages such as non-destructiveness, high efficiency, high precision, and strong adaptability.

[0005] At present, the method of single-shot monitoring of crack parameters based on computer vision has become mature. However, there are still many challenges in carrying out periodic monitoring of crack parameters, including: (1) due to the change of illumination, the grayscale values ​​of crack images collected at different times are different, which can easily lead to inconsistency when extracting crack features; (2) it is difficult to ensure that the camera has the same posture when collecting crack images at different times, which will cause the same crack to present different shapes in the image; (3) it is difficult to ensure that the optical axis of the camera is perpendicular to the surface of the structure being measured, and the difference in spatial resolution in different areas of the image will cause non-negligible measurement errors. Summary of the invention

[0006] The present invention aims at the technical difficulties faced by the periodic monitoring of crack parameters in the development of engineering structures, innovatively integrates methods such as histogram regulation, image calibration and digital image correlation into crack identification and positioning, and proposes a method and system for periodic monitoring of parameters of developing cracks. The present invention solves the technical difficulties of insufficient measurement accuracy caused by illumination changes, spatial resolution differences and crack section positioning errors in periodic monitoring of crack parameters, and can realize efficient, accurate and low-cost periodic monitoring of crack parameters in engineering structures. First, the histogram regulation method is used to adjust the histogram shapes of the same crack images taken at different times to be consistent, so as to facilitate the setting of relevant parameters in crack feature detection, thereby improving the consistency and reliability of the detection results; then, in order to solve the spatial resolution difference caused by the camera posture deviation when collecting crack images at different times, the calibration plate is placed in the crack area when collecting crack images, and the geometric characteristics and scale information of the crack are restored by affine correction and metric correction methods; finally, in order to determine the exact position of the specified section of the same crack, the digital image correlation method is used to search and locate the specified section of the crack in the sequence image. The present invention can provide accurate data support for the growth and evolution of cracks and the health assessment of engineering structures.

[0007] The specific technical solutions of the present invention are as follows:

[0008] 1. A method for periodic monitoring of parameters of developing fractures

[0009] Step 1: Use the camera to collect complete crack images and calibration plate images corresponding to different periods. The camera posture is the same when collecting the complete crack image and calibration plate image in each period; and the complete crack image collected in the initial period is used as the reference crack image;

[0010] Step 2: Generate the histogram of the reference crack image and the histogram corresponding to the complete crack image of other periods, and use the histogram specification method to adjust the histogram shape of other periods so that the histogram shape of other periods is the same as the histogram shape of the reference crack image, thereby obtaining the crack standard image corresponding to all periods;

[0011] Step 3: Combine the calibration plate images of each cycle, use the affine correction and metric correction methods to correct the spatial resolution difference caused by the camera posture deviation in the crack standard image of the current cycle, and obtain the corresponding crack correction image. After traversing and processing the crack standard images of all cycles, obtain the crack correction images corresponding to all cycles;

[0012] Step 4: Calculate the crack width, length and area of ​​the same crack at the preset cross section based on the crack correction diagrams of all cycles.

[0013] The step 1 is specifically as follows:

[0014] Step 1.1: In each cycle, adjust the camera field of view until the camera field of view covers the entire crack, and then acquire a complete crack image; place the calibration plate in the crack area, and acquire the calibration plate image with the same camera posture as the crack image, so as to obtain the complete crack image and calibration plate image of the current cycle;

[0015] Step 1.2: Repeat step 1.1 at different cycles to acquire the complete crack image and calibration plate image corresponding to each cycle, and use the complete crack image acquired in the first cycle as the reference crack image.

[0016] The step 3 is specifically as follows:

[0017] Step 3.1: According to the line parallel property of the calibration plate image, the calibration plate image of each period is used to calculate the affine transformation matrix of the period; then, the calibration plate image of the current period is affine corrected based on the affine transformation matrix to obtain the calibration plate image after affine correction;

[0018] Step 3.2: According to the orthogonality of the calibration plate image, the affine-corrected calibration plate image of the current period is used to calculate the metric transformation matrix of the current period; then, based on the affine transformation matrix and the metric transformation matrix, the crack standard image of the current period is affine-corrected and metric-corrected in turn to obtain the crack correction image of the current period;

[0019] Step 3.3: Repeat steps 3.1 and 3.2 to traverse the crack standard maps of the remaining cycles, thereby obtaining the crack correction maps corresponding to all cycles.

[0020] In step 3.1, according to the line parallel property of the calibration plate image, the calibration plate image of each period is used to calculate the affine transformation matrix of the period, specifically:

[0021] First, a square of the calibration plate is determined and recorded as the calibration quadrilateral, which is mapped to a distorted quadrilateral in each calibration plate image. Each set of parallel sides in the calibration quadrilateral forms an intersection in the distorted quadrilateral. The homogeneous coordinates of the two intersections formed by two sets of parallel sides in the distorted quadrilateral are expressed as (x 1 , y 1 , 1) and ( x 2 , y 2 , 1), the line connecting the two intersection points is recorded as the hidden line l ,satisfy l =( l 1 , l 2 , l 3 ) T ,in l 1 = y 1 - y 2 , l 2 = x 2 - x 1 , l 3 = x 1 y 2 - x 2 y 1 , T is transposed; then based on the hidden line l Generate an affine transformation matrix H AR , satisfying the following formula:

[0022] .

[0023] In step 3.2, according to the orthogonality of the calibration plate image, the affine-corrected calibration plate image of the current period is used to calculate the metric transformation matrix of the period, specifically:

[0024] First, determine the two sets of orthogonal lines in the calibration plate and record them as two sets of marked orthogonal lines. The lines in the two sets of marked orthogonal lines are not parallel to each other. Then determine the corresponding position of each line in the two sets of marked orthogonal lines in the calibration plate image and record them as m , n , A 1 and B 2 , m =( m 1 , m 2 , m3 ), n =( n 1 , n 2 , n 3 ), A 1 =( A 1 1 , A 1 2 , A 1 3 )and B 2 =( B 2 1 , B 2 2 , B 2 3 ), in the calibration plate m ⊥ n and A 1 ⊥ B 2 , satisfying the following formula:

[0025]

[0026] Will KK T The intermediate matrix S , for the intermediate matrix S After decomposition, we obtain an orthogonal matrix U and the diagonal matrix D ,satisfy S= UDU T , and then calculate the first parameter matrix K ,satisfy K = UD 0.5 U T , thus obtaining the metric transformation matrix H MR .

[0027] The step 4 is specifically as follows:

[0028] Step 4.1: After preprocessing the crack correction images of all cycles and detecting the crack edges, the crack edges corresponding to all cycles are obtained;

[0029] Step 4.2: After skeleton extraction of the crack edges corresponding to all cycles, the crack skeletons corresponding to all cycles are obtained;

[0030] Step 4.3: Based on the crack edges corresponding to all cycles, the crack width, length and area of ​​the crack skeleton corresponding to all cycles at the preset cross section are calculated respectively.

[0031] In step 4.1, the preprocessing includes image denoising and threshold segmentation.

[0032] In step 4.3, the crack width calculation process at the preset cross section in each cycle is as follows:

[0033] Firstly, based on the image features at the preset cross section of the crack skeleton of the initial period, the position of the preset cross section in the crack skeleton of other periods is located by using the digital image correlation method;

[0034] Next, select a point in the crack skeleton at the preset section P , with a point P Select 2 along the skeleton line as the center p points and fit the crack skeleton curve of the area based on the selected points, calculate the curvature of the crack skeleton curve of the area, and if it is greater than the curvature threshold, re-fit it using the B-spline curve, otherwise re-fit it using the least squares method to obtain the final fitting skeleton curve; the distance between the two intersection points of the fitting skeleton curve and the crack edge of the current period is recorded as the crack pixel width of the preset section, and then the actual crack width of the preset section is obtained.

[0035] In step 4.3, the crack length at the preset cross section in each period is obtained by sequentially calculating the distances between two adjacent pixel points on the crack skeleton and summing them up.

[0036] 2. A parameter periodic monitoring system for developing fractures

[0037] An image acquisition unit, used to acquire complete crack images and calibration plate images corresponding to different periods;

[0038] A crack standard image generating unit is used to generate a histogram corresponding to a reference crack image and a complete crack image of other periods, and to adjust the histogram shape of other periods by using a histogram specification method so that the histogram shape of other periods is the same as the histogram shape corresponding to the reference crack image, thereby obtaining crack standard images corresponding to all periods;

[0039] An image correction unit is used to combine the calibration plate image of each period, and use affine correction and metric correction methods to correct the spatial resolution difference caused by the camera posture deviation in the crack standard image of the current period, obtain the corresponding crack correction image, and obtain the crack correction images corresponding to all periods after traversing and processing the crack standard images of all periods;

[0040] The crack parameter generating unit is used to calculate the crack width, length and area of ​​the same crack at a preset cross section according to the crack correction diagrams of all cycles.

[0041] 3. A computer device

[0042] The computer device comprises a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the method for periodic parameter monitoring of developing fractures when executing the computer program.

[0043] 4. A computer-readable storage medium

[0044] The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method for periodic parameter monitoring of developing fractures are implemented.

[0045] 5. A computer program product

[0046] The computer program product comprises a computer program / instruction, which, when executed by a processor, implements the steps of the method for periodic parameter monitoring of developing fractures.

[0047] The beneficial effects of the present invention are:

[0048] The present invention can improve the accuracy and reliability of parameter calculation of cracks in the development of engineering structures. In addition, compared with the method of long-term monitoring and analysis of crack evolution laws, the high-precision periodic monitoring method of crack parameters proposed by the present invention has obvious improvements in calculation efficiency, monitoring cost, and operational convenience. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 It is a schematic diagram of the method flow of the present invention.

[0050] Figure 2 This is the checkerboard calibration plate used in the present invention.

[0051] Figure 3 Schematic diagram of the difference in spatial resolution of images corrected by affine correction and metric correction; (a) is a schematic diagram of the difference in spatial resolution of images corrected by affine correction before and after, and (b) is a schematic diagram of the difference in spatial resolution of images corrected by metric correction before and after.

[0052] Figure 4Schematic diagram of the crack area in this embodiment.

[0053] Figure 5 Match the schematic diagram to the specified cross section of the crack.

[0054] Figure 6 Schematic diagram of crack width calculation based on skeleton line.

[0055] Figure 7 Schematic diagram of crack length calculation based on Euclidean distance.

[0056] Figure 8 The calibration images and crack images are collected at regular intervals in this embodiment; (a) is the calibration image collected for the first time, (b) is the complete crack image collected for the first time, (c) is the calibration image collected for the second time, and (d) is the complete crack image collected for the second time.

[0057] Fig. 9 Schematic diagram of the crack image preprocessing and skeleton extraction process in the embodiment.

[0058] Fig.10 Schematic diagram of the difference in width of the crack specified cross section measured on October 20, 2023 and March 16, 2024 in the embodiment. DETAILED DESCRIPTION

[0059] In order to enable those skilled in the art to better understand the present invention, and thus to more clearly define the scope of the present invention, the present invention is described in detail with respect to some specific embodiments of the present invention. It should be noted that the following are only some specific implementation methods of the present invention, which are only part of the embodiments of the present invention, wherein the specific and direct description of the relevant structure is only for the convenience of understanding the present invention, and each specific feature does not naturally and directly limit the scope of implementation of the present invention. The conventional selection and replacement made by those skilled in the art under the guidance of the present invention, as well as the reasonable arrangement and combination of several technical features under the guidance of the present invention, should be regarded as within the scope of the present invention.

[0060] The present invention proposes a method and system for high-precision crack parameter measurement for periodic monitoring. The present invention collects images of the same crack of an engineering structure at certain intervals to calculate crack parameters and analyze their evolution laws. A calibration plate is placed in the crack area each time a crack image is collected, and the spatial resolution difference caused by the camera posture deviation is corrected by affine rectification (AR) and metric rectification (MR) methods. The histogram shape of images of the same crack at different times is adjusted to be consistent using the histogram specification (HS) method to improve the consistency and reliability of crack edge detection. The digital image correlation method is used to search and locate the specified crack section to determine the precise position of the same crack specified section in different images. The present invention does not require the construction of fixed observation piers or high-precision sensors, and only requires a camera and a calibration plate to achieve efficient and accurate calculation of crack parameters and analysis of their evolution laws. The method flow chart is shown below. Figure 1 As shown, the specific process includes the following:

[0061] Step 1: Use the camera to collect complete crack images corresponding to different periods ( Img 1 OG , Img 2 OG , …, Img n OG ) and the calibration plate image ( img 1 OG , img 2 OG , …, img n OG ), the camera posture is the same when collecting the complete crack image and the calibration plate image in each cycle; and the complete crack image collected in the initial cycle (i.e., the first cycle) Img 1 OG as the reference crack image; the complete crack image and calibration plate image collected for the first and second times are shown in Figure 8 (a) Figure 8 (b) Figure 8 (c) and Figure 8 as shown in (d).

[0062] Step 1 is as follows:

[0063] Step 1.1: In each cycle, adjust the camera field of view until the camera field of view covers the entire crack, and then collect a complete image of the crack; place a calibration plate in the crack area. The calibration plate is a black and white chessboard, such as Figure 2 As shown. The calibration plate image is acquired with the same camera posture as the crack image (that is, the camera does not move), so as to obtain the complete crack image and calibration plate image of the current cycle;

[0064] Step 1.2: Repeat step 1.1 at different cycles to acquire the complete crack image and calibration plate image corresponding to each cycle, and use the complete crack image acquired in the first cycle as the reference crack image. The camera poses in different cycles are not necessarily the same.

[0065] Step 2: Generate the histograms corresponding to the reference crack image and the complete crack images of other periods, and use the histogram specification method to adjust the histogram shapes of other periods so that the histogram shapes of other periods are the same as the histogram shapes corresponding to the reference crack image, thereby obtaining the crack standard images corresponding to all periods ( Img 1 HS , Img 2 HS , …, Img n HS ), that is, the corresponding histogram shapes of the crack standard images of different periods are the same, and the crack standard image of the initial period is the reference crack image; this step is conducive to the subsequent setting of relevant parameters when using image denoising and threshold segmentation methods to identify cracks, so as to improve the consistency and reliability of the detection results.

[0066] Step 3: Combine the calibration plate images of each cycle, use the affine correction and metric correction methods to correct the spatial resolution difference caused by the camera posture deviation in the crack standard image of the current cycle, and obtain the corresponding crack correction image. After traversing and processing the crack standard images of all cycles, obtain the crack correction images corresponding to all cycles;

[0067] Step 3 is as follows:

[0068] Step 3.1: According to the parallel properties of the lines in the calibration plate image, use the calibration plate image of each period to calculate the affine transformation matrix of the period; then perform affine correction on the calibration plate image of the current period based on the affine transformation matrix to obtain the affine-corrected calibration plate image; the affine-corrected image only restores the parallel properties of the parallel lines in the image, and the angle still presents a distorted state that does not conform to the real scene, and it still needs to be metrically corrected.

[0069] Among them, according to the parallel property of the lines of the calibration plate image, the calibration plate image of each period is used to calculate the affine transformation matrix of the period, specifically:

[0070] First, a square of the calibration plate is determined and recorded as the calibration quadrilateral, which is mapped to a distorted quadrilateral in each calibration plate image. Each set of parallel sides in the calibration quadrilateral forms an intersection in the distorted quadrilateral. The homogeneous coordinates of the two intersections formed by two sets of parallel sides in the distorted quadrilateral are expressed as ( x 1 , y 1 , 1) and ( x 2 , y 2 , 1), the line connecting the two intersection points is recorded as the hidden line l =( l 1 , l 2 , l 3 ) T ,in l 1 = y 1 - y 2 , l 2 = x 2 - x 1 , l 3 = x 1 y 2 - x 2 y 1 , T is transposed; then based on the hidden line l Generate an affine transformation matrix H AR , satisfying the following formula:

[0071] .

[0072] Take the first set of crack images Img 1 OG And the calibration plate image img 1 OG For example, the collected calibration plate image img 1 OG The schematic diagram is as follows Figure 3 It contains two sets of parallel lines of the real scene A 1 ∥A 2 and B 1 ∥ B 2 ,and A 1 ⊥ B 1 , A 2 ⊥ B 2 . A 1 , A 2 Intersection point p 1 , B 1 , B 2 Intersection point p 2 . p 1 and p 2 The line connecting the l =( l 1 , l 2 , l 3 ) T , l 1 , l 2 , l 3 can be p 1 and p 2 The coordinates of are determined by the affine transformation matrix H 1 AR Can map hidden lines to straight lines at infinity l ∞ =(0,0,1) T In, that is ( H 1 AR )^(-T)= l ∞ , ^ represents the power. In the obtained affine transformation matrix H 1 AR After that, the calibration plate image img 1 OG Perform affine correction to obtain the calibration plate image after affine correction img 1AR ,like Figure 3 As shown in (a).

[0073] Step 3.2: According to the orthogonality of the calibration plate image, the affine-corrected calibration plate image of the current period is used to calculate the metric transformation matrix of the period; then, based on the affine transformation matrix and the metric transformation matrix, the crack standard image of the current period is affine-corrected and metric-corrected in turn to obtain the crack correction image of the current period, thereby restoring the true geometric information and morphology of the crack;

[0074] According to the orthogonality of the calibration plate image, the affine-corrected calibration plate image of the current period is used to calculate the metric transformation matrix of the period, specifically:

[0075] First, determine the two sets of orthogonal lines in the real scene in the calibration plate and record them as two sets of marked orthogonal lines. The lines in the two sets of marked orthogonal lines are not parallel to each other. Then determine the corresponding position of each line in the two sets of marked orthogonal lines in the calibration plate image and record them as m , n , A 1 and B 2 ,like Figure 3 As shown in (b), m =( m 1 , m 2 , m 3 ), n =( n 1 , n 2 , n 3 ), A 1 =( A 1 1 , A 1 2 , A 1 3 )and B 2 =( B 2 1 , B 2 2 , B 2 3 ), in the calibration plate m ⊥ n andA 1 ⊥ B 2 , satisfying the following formula:

[0076]

[0077] Will KK T The intermediate matrix S , for the intermediate matrix S After decomposition, we obtain an orthogonal matrix U and the diagonal matrix D ,satisfy S = UDU T , and then calculate the first parameter matrix K ,satisfy K = UD 0.5 U T , thus obtaining the metric transformation matrix H MR .

[0078] Step 3.3: Repeat steps 3.1 and 3.2 to traverse the crack standard maps of the remaining cycles, thereby obtaining the crack correction maps corresponding to all cycles.

[0079] Step 4: Calculate the crack width, length and area of ​​the same crack at the preset cross section based on the crack correction diagrams of all cycles.

[0080] Step 4 is as follows:

[0081] Step 4.1: Fig. 9 As shown, after preprocessing the crack correction images of all cycles and detecting the crack edges, the crack edges corresponding to all cycles are obtained; wherein the preprocessing includes image denoising and threshold segmentation.

[0082] Step 4.2: After skeleton extraction of the crack edges corresponding to all cycles, the crack skeletons corresponding to all cycles are obtained; the crack skeleton is the center line that can represent the most basic form of the crack, so as to more accurately represent and analyze the geometric characteristic structure of the crack, such as Figure 4 shown.

[0083] Step 4.3: Based on the crack edges corresponding to all cycles, the crack width, length and area of ​​the crack skeleton corresponding to all cycles at the preset cross section are calculated respectively.

[0084] The calculation process of the crack width at the preset cross section for each cycle is as follows:

[0085] Firstly, based on the image features at the preset cross section of the crack skeleton of the initial period, the digital image correlation method (DIC) is used to locate the position of the preset cross section in the crack skeleton of other periods.

[0086] by Figure 4 For example, the crack skeleton in the initial cycle Img 1 MR Select the specified section of the crack and intersect the skeleton line at the point P , using digital image correlation (DIC) to locate P Crack skeleton in other cycles Img 2 MR , …, Img n MR The specific steps are as follows:

[0087] Assume the crack skeleton of the initial period Img 1 MR f(x ij ), Img n MR (n=2,…,n) is g(x ij ), in order to determine P Crack skeleton in other cycles Img 2 MR , …, Img n MR The position in the P The crack skeleton centered at the initial cycle Img 1 MR Select M × M As a reference sub-area, Figure 5 By converting the texture information of the reference sub-area into grayscale information, and Img 2 MR , …, Img n MR The most similar target sub-region is searched in the search result. The similarity between the reference sub-region and the target sub-region can be represented by a correlation coefficient. The present invention uses a zero-mean normalized least square distance correlation function to calculate the correlation between the two.

[0088] Next, select a point in the crack skeleton at the preset section P , with a point P Select 2 along the skeleton line as the centerp Points are selected and the crack skeleton curve of the region is obtained by fitting based on the selected points. The curvature of the crack skeleton curve of the region is calculated. If it is greater than the curvature threshold, the curvature threshold is 0.1, then the B-spline curve is used to refit. Otherwise, the least square method is used to refit to obtain the final fitting skeleton curve. The distance between the two intersection points of the fitting skeleton curve and the crack edge of the current cycle is recorded as the crack pixel width of the preset section, such as Figure 6 The crack width obtained at this time is measured in pixels, and it is still necessary to use the scale information of the calibration plate to convert it into the actual physical size. D , and then obtain the actual width of the crack of the preset cross section. Therefore, for cracks with slight or moderate curvature, the present invention uses the least squares method to fit the quadratic polynomial, which can better approximate the actual crack curvature. For cracks with local large curvature, the B-spline curve is used to achieve a better fitting effect.

[0089] The crack length at the preset cross section in each cycle is obtained by sequentially calculating the distances between two adjacent pixel points on the crack skeleton and summing them up. Figure 7 As shown, assuming that the first k and k The image coordinates of the +1 pixel are (x k , y k ) and (x k+1 , y k+1 ), the Euclidean distance between them l k It can be expressed as:

[0090] l k = c k ((x k+1 -x k )+ (y k+1 ,-y k ))^(1 / 2)

[0091] Among them, c k Indicates k and k +1 pixel corresponding to the average physical size, the calculation formula for the length L of the entire crack can be written as:

[0092] L=∑ l k , (k=1,2,…n)

[0093] The crack area at the preset cross section in each cycle is calculated based on the region growing method. Figure 4In the crack region shown, a pixel is selected as the starting point, and all cracks connected to the starting pixel are connected to gradually grow a crack region, and the area of ​​the region is the crack area.

[0094] Crack skeleton according to all cycles Img 2 MR , Img 2 MR , …, Img n M The crack width, length and area in the image can be subtracted from the corresponding values ​​to obtain the development trend of the cracks, and then analyze their evolution law.

[0095] In this example, the lengths of the same cracks collected about 150 days apart were 356.58 mm and 356.42 mm, respectively, and the areas were 1731.25 m 2 and 1737.37 m 2 , and the width of the specified section are 5.04 mm and 5.11 mm respectively. In addition, this example counts the width changes of 35 sections of the crack at intervals of about 150 days, such as Fig.10 As shown. It can be seen that the high-precision crack parameter periodic monitoring method and system proposed in the present invention solves the technical difficulties of insufficient measurement accuracy caused by illumination changes, spatial resolution differences and crack section positioning errors, and can provide accurate data support for the exploration of crack evolution laws and structural health assessment. At the same time, its implementation process is relatively simple and easy.

[0096] The present invention also proposes a parameter periodic monitoring system for developing cracks, comprising:

[0097] An image acquisition unit, used to acquire complete crack images and calibration plate images corresponding to different periods;

[0098] A crack standard image generating unit is used to generate a histogram corresponding to a reference crack image and a complete crack image of other periods, and to adjust the histogram shape of other periods by using a histogram specification method so that the histogram shape of other periods is the same as the histogram shape corresponding to the reference crack image, thereby obtaining crack standard images corresponding to all periods;

[0099] An image correction unit is used to combine the calibration plate image of each period, and use affine correction and metric correction methods to correct the spatial resolution difference caused by the camera posture deviation in the crack standard image of the current period, obtain the corresponding crack correction image, and obtain the crack correction images corresponding to all periods after traversing and processing the crack standard images of all periods;

[0100] The crack parameter generating unit is used to calculate the crack width, length and area of ​​the same crack at a preset cross section according to the crack correction diagrams of all cycles.

[0101] The present invention also proposes a computer device, which includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, the steps of a parameter periodic monitoring method for developing fractures are implemented.

[0102] The present invention also proposes a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of a method for periodic parameter monitoring of developing fractures are implemented.

[0103] The present invention also proposes a computer program product, which includes a computer program / instruction. When the computer program / instruction is executed by a processor, the steps of a method for periodic parameter monitoring of developing fractures are implemented.

Claims

1. A method for periodic monitoring of parameters of developing cracks, characterized in that: The following steps are involved: Step 1: Use the camera to collect complete crack images and calibration plate images corresponding to different periods. The camera posture is the same when collecting the complete crack image and calibration plate image in each period; and the complete crack image collected in the initial period is used as the reference crack image; Step 2: Generate the histogram of the reference crack image and the histogram corresponding to the complete crack image of other periods, and use the histogram specification method to adjust the histogram shape of other periods so that the histogram shape of other periods is the same as the histogram shape of the reference crack image, thereby obtaining the crack standard image corresponding to all periods; Step 3: Combine the calibration plate images of each cycle, use the affine correction and metric correction methods to correct the spatial resolution difference caused by the camera posture deviation in the crack standard image of the current cycle, and obtain the corresponding crack correction image. After traversing and processing the crack standard images of all cycles, obtain the crack correction images corresponding to all cycles; The step 3 is specifically as follows: Step 3.1: According to the line parallel property of the calibration plate image, the calibration plate image of each period is used to calculate the affine transformation matrix of the period; then, the calibration plate image of the current period is affine corrected based on the affine transformation matrix to obtain the calibration plate image after affine correction; In step 3.1, according to the line parallel property of the calibration plate image, the calibration plate image of each period is used to calculate the affine transformation matrix of the period, specifically: First, a square of the calibration plate is determined and recorded as the calibration quadrilateral, which is mapped to a distorted quadrilateral in each calibration plate image. Each set of parallel sides in the calibration quadrilateral forms an intersection in the distorted quadrilateral. The homogeneous coordinates of the two intersections formed by two sets of parallel sides in the distorted quadrilateral are expressed as ( x 1, y 1, 1) and ( x 2, y 2, 1), the line connecting the two intersection points is recorded as the hidden line l ,satisfy l =( l 1, l 2, l 3) T ,in l 1= y 1- y 2, l 2= x 2- x 1, l 3= x 1 y 2- x 2 y 1, T is transposed; then based on the hidden line l Generate an affine transformation matrix H AR , satisfying the following formula: ； Step 3.2: According to the orthogonality of the calibration plate image, the affine-corrected calibration plate image of the current period is used to calculate the metric transformation matrix of the current period; then, based on the affine transformation matrix and the metric transformation matrix, the crack standard image of the current period is affine-corrected and metric-corrected in turn to obtain the crack correction image of the current period; Step 3.3: Repeat steps 3.1 and 3.2 to traverse and process the crack standard maps of the remaining cycles, thereby obtaining the crack correction maps corresponding to all cycles; Step 4: Calculate the crack width, length and area of ​​the same crack at the preset cross section based on the crack correction diagrams of all cycles.

2. A parameter periodic monitoring method for developing cracks according to claim 1, characterized in that: The step 1 is specifically as follows: Step 1.1: In each cycle, adjust the camera field of view until the camera field of view covers the entire crack, and then acquire a complete crack image; place the calibration plate in the crack area, and acquire the calibration plate image with the same camera posture as the crack image, so as to obtain the complete crack image and calibration plate image of the current cycle; Step 1.2: Repeat step 1.1 at different cycles to acquire the complete crack image and calibration plate image corresponding to each cycle, and use the complete crack image acquired in the first cycle as the reference crack image.

3. A parameter periodic monitoring method for developing cracks according to claim 1, characterized in that: In step 3.2, according to the orthogonality of the calibration plate image, the affine-corrected calibration plate image of the current period is used to calculate the metric transformation matrix of the period, specifically: First, determine the two sets of orthogonal lines in the calibration plate and record them as two sets of marked orthogonal lines. The lines in the two sets of marked orthogonal lines are not parallel to each other. Then determine the corresponding position of each line in the two sets of marked orthogonal lines in the calibration plate image and record them as m , n , A 1 and B 2 , m =( m 1, m 2, m 3), n =( n 1, n 2, n 3), A 1 =( A 1 1, A 1 2, A 1 3) and B 2 =( B 2 1, B 2 2, B 2 3) In the calibration plate, m ⊥ n and A 1 ⊥ B 2 , satisfying the following formula: H MR = Will KK T The intermediate matrix S , for the intermediate matrix S After decomposition, we obtain an orthogonal matrix U and the diagonal matrix D ,satisfy S = UDU T , and then calculate the first parameter matrix K ,satisfy K = UD 0.5 U T , thus obtaining the metric transformation matrix H MR .

4. A parameter periodic monitoring method for developing cracks according to claim 1, characterized in that: The step 4 is specifically as follows: Step 4.1: After preprocessing the crack correction images of all cycles and detecting the crack edges, the crack edges corresponding to all cycles are obtained; Step 4.2: After skeleton extraction of the crack edges corresponding to all cycles, the crack skeletons corresponding to all cycles are obtained; Step 4.3: Based on the crack edges corresponding to all cycles, the crack width, length and area of ​​the crack skeleton corresponding to all cycles at the preset cross section are calculated respectively.

5. A parameter periodic monitoring method for developing cracks according to claim 4, characterized in that: In step 4.1, the preprocessing includes image denoising and threshold segmentation.

6. A method for periodic monitoring of parameters of developing fractures according to claim 4, characterized in that: In step 4.3, the crack width calculation process at the preset cross section in each cycle is as follows: Firstly, based on the image features at the preset cross section of the crack skeleton of the initial period, the position of the preset cross section in the crack skeleton of other periods is located by using the digital image correlation method; Next, a point P in the crack skeleton at the preset section is selected, and a total of 2p points are selected along the skeleton line with point P as the center, and the crack skeleton line curve of the preset section is obtained by fitting based on the selected points. The curvature of the crack skeleton curve of the preset section is calculated. If it is greater than the curvature threshold, the B-spline curve is used to re-fit, otherwise the least squares method is used to re-fit to obtain the final fitting skeleton curve; the distance between the two intersection points of the fitting skeleton curve and the crack edge of the current period is recorded as the crack pixel width of the preset section, and then the actual crack width of the preset section is obtained.

7. A method for periodic monitoring of parameters of developing fractures according to claim 4, characterized in that: In step 4.3, the crack length at the preset cross section in each period is obtained by sequentially calculating the distances between two adjacent pixel points on the crack skeleton and summing them up.

8. A parameter periodic monitoring system for developing fractures, characterized in that: include: An image acquisition unit, used to acquire complete crack images and calibration plate images corresponding to different periods; A crack standard image generating unit is used to generate a histogram corresponding to a reference crack image and a complete crack image of other periods, and to adjust the histogram shape of other periods by using a histogram specification method so that the histogram shape of other periods is the same as the histogram shape corresponding to the reference crack image, thereby obtaining crack standard images corresponding to all periods; An image correction unit is used to combine the calibration plate image of each period, and use affine correction and metric correction methods to correct the spatial resolution difference caused by the camera posture deviation in the crack standard image of the current period, obtain the corresponding crack correction image, and obtain the crack correction images corresponding to all periods after traversing and processing the crack standard images of all periods; In the image correction unit, the specific processing steps are as follows: S1: According to the line parallel property of the calibration plate image, the calibration plate image of each period is used to calculate the affine transformation matrix of the period; then, the calibration plate image of the current period is affine corrected based on the affine transformation matrix to obtain the calibration plate image after affine correction; According to the parallel property of the lines of the calibration plate image, the calibration plate image of each period is used to calculate the affine transformation matrix of the period, specifically: First, a square of the calibration plate is determined and recorded as the calibration quadrilateral, which is mapped to a distorted quadrilateral in each calibration plate image. Each set of parallel sides in the calibration quadrilateral forms an intersection in the distorted quadrilateral. The homogeneous coordinates of the two intersections formed by two sets of parallel sides in the distorted quadrilateral are expressed as ( x 1, y 1, 1) and ( x 2, y 2, 1), the line connecting the two intersection points is recorded as the hidden line l ,satisfy l =( l 1, l 2, l 3) T ,in l 1= y 1- y 2, l 2= x 2- x 1, l 3= x 1 y 2- x 2 y 1, T is transposed; then based on the hidden line l Generate an affine transformation matrix H AR , satisfying the following formula: ； S2: According to the orthogonality of the calibration plate image, the metric transformation matrix of the period is calculated using the affine-corrected calibration plate image of the current period; then, based on the affine transformation matrix and the metric transformation matrix, affine correction and metric correction are performed on the crack standard image of the current period in turn to obtain the crack correction image of the current period; S3: repeat S1 and S2, traverse and process the crack standard maps of the remaining cycles, so as to obtain the crack correction maps corresponding to all cycles; The crack parameter generating unit is used to calculate the crack width, length and area of ​​the same crack at a preset cross section according to the crack correction diagrams of all cycles.

Images