A crack monitoring method

Abstract

This invention provides a crack monitoring method, comprising: Step S1, placing a pair of circular targets on both sides of the crack, wherein each circular target includes a coding ring and a target center circle located inside the coding ring; Step S2, acquiring a crack image containing the crack and the pair of circular targets captured by a camera; correcting the crack image and converting it into a grayscale image; performing edge detection on the grayscale image, and then performing ellipse search based on the detected contours; Step S3, detecting the ellipse that is the target center circle of the circular targets among the identified ellipses; Step S4, for the two identified circular targets, performing perspective transformation on the target area based on the common tangent of the ellipses to obtain an orthophoto image of the crack, calculating the actual distance between the target centers of the two circular targets, and realizing crack monitoring. By automatically calculating the crack width using paired circular targets, changes in cracks in buildings can be accurately monitored, and the monitoring accuracy can reach the sub-millimeter level when taking photos within 1 meter.

Description

Technical Field

[0001] This invention relates to the fields of surveying, photogrammetry, and structural safety technology, and specifically to a crack monitoring method. Background Technology

[0002] During the construction and operation phases of buildings and structures, deformation inevitably occurs due to the structure's own weight, soil unloading, and structural loads, leading to cracks on the structural surface. Over time, existing small cracks may widen or even develop into new cracks, threatening structural health. Therefore, it is necessary to monitor cracks in buildings and structures, observe major or significantly changing cracks, take current photographs and describe their shapes, and establish a crack status archive for the buildings and structures.

[0003] Currently, methods for monitoring cracks in buildings and structures mainly fall into two categories: The first method is manual monitoring, which involves placing measuring points on both sides of the crack. These points are typically marked with paint, paper, or steel nails, with a pen marking the center of each point to ensure consistent measurement. During each monitoring session, a total station is used to measure the coordinates of the marked points on both sides of the crack, and the difference in coordinates is used to analyze changes in crack width. The second method is automated monitoring, commonly using a string-type displacement gauge. This method offers high accuracy, reaching sub-millimeter levels, but it involves complex on-site installation and maintenance, and incurs high hardware costs.

[0004] Currently, there are also studies on crack monitoring using image methods. Generally, markers are placed on both sides of the crack, and then the crack is identified and monitored using deep learning and other methods. This method is simple to deploy on site, but since deep learning requires a large number of samples for learning, it is difficult to implement when there is no sample data or the sample data is small.

[0005] In conclusion, there is an urgent need for a crack monitoring method to address the problems existing in current technologies. Summary of the Invention

[0006] The purpose of this invention is to provide a crack monitoring method, which aims to overcome the drawbacks of existing image-based crack monitoring methods. The specific technical solution is as follows:

[0007] A crack monitoring method includes the following steps:

[0008] Step S1: Place a pair of circular targets on both sides of the crack. The circular targets include a coding ring and a target center circle located inside the coding ring.

[0009] Step S2: Acquire a crack image captured by the camera, containing the crack and a pair of circular targets; correct the crack image and convert it into a grayscale image; perform edge detection on the grayscale image, and then perform ellipse search based on the detected contours;

[0010] Step S3: Among the ellipses identified in step S2, detect the ellipse with the center circle of the circular target.

[0011] Step S4: For the two identified circular targets, perform perspective transformation on the target area based on the common tangent of the ellipse to obtain the orthophoto image of the crack, calculate the actual distance between the centers of the two circular targets, and realize the monitoring of the crack.

[0012] In the preferred embodiment of the above technical solution, the coding ring has a total of 360°, with each N degrees being one coding bit, for a total of 360 / N coding bits; the value of N is in the range of [10, 60], and N is divisible by 360, with the unit of N being °.

[0013] In the preferred embodiment of the above technical solution, in step S2, if a contour satisfies expression 1), it is considered an ellipse:

[0014]

[0015] Where ptsNum is the number of points contained in the contour, polyDP is the number of vertices of the convex polygon fitted by the contour, contourArea is the area of ​​the contour, and rectArea is the area of ​​the minimum bounding rectangle of the contour.

[0016] In a preferred embodiment of the above technical solutions, step S3 includes the following steps:

[0017] Step S3.1: Binarize the grayscale image according to expression 2):

[0018] threshold=(minGray+aveGray) / rate 2),

[0019] Where threshold is the grayscale threshold for binarization; minGray is the minimum grayscale value of the grayscale image; aveGray is the average grayscale value of the grayscale image; rate is the ratio, and its value range is [2, 4].

[0020] Step S3.2: Find the four vertices of the minimum bounding rectangle of the ellipse based on the parameters of the ellipse, transform this minimum bounding rectangle into a square and solve for the affine transformation matrix, and perform an affine transformation on the binarized image based on the affine transformation matrix. After the affine transformation, the ellipse is transformed into a circle; the center of the circle after the affine transformation is (circleX, circleY), and the radius is radius.

[0021] Step S3.3: If the ellipse satisfies expression 3), then it is a potential target circle.

[0022]

[0023] Wherein, innerPixels is the number of pixels inside the target circle, blackPixels is the number of pixels with a value of 0 inside the target circle, outerPixels is the number of pixels in the region between the target circle and the coding ring, whitePixels is the number of pixels with a value of 1 in the region between the target circle and the coding ring, and percent takes values ​​in the range of [0.8, 1].

[0024] Step S3.4, with (circleX, circleY) as the center, A circular ring is formed with radius p. A pixel is taken every ang on the ring, resulting in a total of 360 / ang pixels p. i The subscript i represents the pixel number; ang takes values ​​in the range [1, 5] and is divisible by N, with the unit being °; R2 is the inner radius of the encoding ring, and R3 is the outer radius of the encoding ring;

[0025] Obtain each pixel p in sequence i The corresponding pixel value g i Search for g i The sequence number q corresponding to the first mutation site in the sequence;

[0026] From g i Take from sequence Where k is an integer and its value ranges from 1 to 1. get There are 1 data group; N is the angle corresponding to the coded bit in the coding ring;

[0027] Let count be the number of 0 values ​​in the k-th data group. k The encoded value (bit) of the k-th data group is shown in expression 5):

[0028]

[0029] exist If any bit in the encoded value of any data group contains a negative value, then the ellipse is not the center circle of the circular target; if none of the bits in the encoded value of any data group contain a negative value, then the ellipse is the center circle of the circular target.

[0030] In the preferred embodiment of the above technical solutions, after identifying the ellipse as the target circle, the value of this coded ring is obtained. The binary number corresponding to Code is cyclically shifted to obtain the minimum value minCode, which is used as the encoding value of the corresponding circular target.

[0031] In the preferred embodiment of the above technical solutions, the pixel (x, y) in innerPixels satisfies The pixels (x, y) in outerPixels satisfy the following conditions: R1 is the radius of the target circle.

[0032] In a preferred embodiment of the above technical solutions, step S4 includes the following steps:

[0033] Step 54.1: Perform sub-pixel extraction on the ellipses EllipseA and EllipseB corresponding to the center circles of the two identified circular targets, and obtain the accurate ellipse equations respectively;

[0034] Step S4.2: Obtain the equations of the common tangents between all common tangents between ellipses EllipseA and EllipseB;

[0035] Step S4.3: Project the tangent points of the two external common tangents to the ellipses EllipseA and EllipseB respectively to the minimum bounding rectangles corresponding to the tangent points to construct the perspective transformation matrix M, and perform perspective transformation on the crack image according to the perspective transformation matrix M.

[0036] Step S4.4: Let the ellipses EllipseA and EllipseB after perspective transformation be EllipseA′ and EllipseB′ respectively. Perform affine transformations on ellipses EllipseA′ and EllipseB′ respectively to obtain image A and image B.

[0037] Step S4.5: Calculate the actual distance between the centers of the two circular targets based on images A and B.

[0038] In the preferred embodiment of the above technical solution, in step S4.1, the general equation of the ellipse is obtained based on the identified ellipse; at least 5 points on the ellipse are selected to solve for their respective sub-pixel coordinates; and the accurate ellipse equation is obtained by using the least squares method based on the obtained sub-pixel coordinate values.

[0039] The preferred method for solving sub-pixel coordinates in the above technical solutions is as follows:

[0040] Let the coordinates of a point on the ellipse be (ex, ey). Solve for the unit normal vector of the normal line corresponding to this point as (nX, nY). Take m pixels on each side of the point along the normal line. Denote the pixels on both sides of the normal line and the point (ex, ey) as point pj, where j = -m, -m+1, ..., m.

[0041] Calculate p at each point j gray derivative g′ j , of which 9′ j =|g j -g j-1 |+|g j+1 -g j |;g j It is point p j Pixel values;

[0042] For the obtained g′ j Perform Gaussian curve fitting and find the mean extreme point of the corresponding Gaussian curve. The pixel coordinates corresponding to the mean extreme point are (gx, gy), where (gx, gy) are the sub-pixel coordinates of the point (ex, ey).

[0043]

[0044] The preferred technical solution among the above is:

[0045] In image A, the center coordinates of the circle after EllipseA′ transformation are (circleA_X1, circleA_Y1), the radius is radiusA, and the area of ​​the circle is areaA1. The center coordinates of the circle after EllipseB′ transformation are (circleB_X1, circleB_Y1), and the area of ​​the circle is areaB1. The distance between the centers of the two circular targets in image A is:

[0046]

[0047] In image B, the center coordinates of the circle after EllipseA′ transformation are (circleA_X2, circleA_Y2), and the area of ​​the circle is areaA2. The center coordinates of the circle after EllipseB′ transformation are (circleB_X2, circleB_Y2), the radius is radiusB, and the area of ​​the circle is areaB2. The distance between the centers of the two circular targets in image B is:

[0048]

[0049] The actual distance between the centers of the two circular targets

[0050] in: R1 is the radius of the target circle.

[0051] The application of the technical solution of the present invention has the following beneficial effects:

[0052] The monitoring method of this invention solves the problem of existing technologies that require a large number of samples for learning in crack identification and monitoring through deep learning. By automatically calculating the crack width using paired circular targets, it can accurately monitor changes in cracks in buildings and structures. When taking photographs within 1 meter, the monitoring accuracy can reach the sub-millimeter level. The method of this invention is simple to deploy on-site, low in cost, and ensures high monitoring accuracy, thus improving the automation level of crack monitoring.

[0053] In addition to the objectives, features, and advantages described above, the present invention has other objectives, features, and advantages. The invention will now be described in further detail with reference to the figures. Attached Figure Description

[0054] The accompanying drawings, which form part of this application, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings:

[0055] Figure 1 This is a flowchart of the monitoring method;

[0056] Figure 2 This is a schematic diagram of a circular target;

[0057] Figure 3 This is a schematic diagram of target encoding reading after affine transformation;

[0058] Figure 4 This is a schematic diagram of the common tangent and the point of tangency of an ellipse. Detailed Implementation

[0059] To facilitate understanding of the present invention, a more comprehensive description is provided below, along with preferred embodiments. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided to provide a thorough and complete understanding of the disclosure of the present invention.

[0060] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.

[0061] Example 1:

[0062] A crack monitoring method involves first placing circular targets on both sides of a crack in a building structure, then using a camera to photograph the crack and simultaneously capturing images of the targets on both sides. The captured images are then processed to extract edges, identify the circular targets, and extract sub-pixel edges of the targets. Perspective transformation is performed on the images based on the common tangent of the pair of targets to restore them to orthophotos. Finally, the actual distance between the centers of the pair of targets is calculated to monitor the crack.

[0063] like Figure 1 As shown, the detailed steps of the crack monitoring method in this embodiment are as follows:

[0064] Step S1: Place a pair of circular targets on both sides of the crack (i.e., place circular targets on each side of the crack), wherein each circular target includes a coding ring and a target center circle located inside the coding ring; see [link to previous step]. Figure 2 The radius of the target circle is R1, the radius of the inner circle of the coding ring is R2, and the radius of the outer circle of the coding ring is R3.

[0065] Preferably, the coding ring has a total of 360°, with each N degrees being one coding bit, for a total of 360 / N coding bits. The value of N is in the range of [10, 60], and the unit of N is ° and N is divisible by 360.

[0066] Specifically, in this embodiment, N is set to 30, meaning that each 30° is a coding bit, for a total of 12 coding bits, resulting in 351 unordered coding values. The circular targets are sorted in ascending order of coding values ​​from [1, 2047], with each pair of targets forming a pair, resulting in a total of 175 coding pairs.

[0067] Step S2: Acquire a crack image captured by the camera, containing the crack and a pair of circular targets; correct the crack image and convert it into a grayscale image; perform edge detection on the grayscale image, and then perform ellipse search based on the detected contours;

[0068] Specifically, if a contour satisfies expression 1), it is considered an ellipse:

[0069]

[0070] Where ptsNum is the number of points contained in the contour, polyDP is the number of vertices of the convex polygon fitted by the contour, contourArea is the area of ​​the contour, and rectArea is the area of ​​the minimum bounding rectangle of the contour.

[0071] The geometric center of the ellipse that satisfies expression 1) is (centreX, centerY), the minor axis is a, the major axis is b, and the angle between the minor axis and the horizontal axis is α (clockwise is positive).

[0072] Preferably, the crack image correction specifically involves: calibrating the camera using a checkerboard target based on the Zhang Zhengyou calibration method, obtaining the camera's intrinsic parameter matrix mtx, distortion coefficient dist, rotation matrix rvecs, and translation vector tvecs, and correcting the crack image based on the intrinsic parameter matrix mtx, distortion coefficient dist, rotation matrix rvecs, and translation vector tvecs.

[0073] Step S3: Among the ellipses identified in step S2, detect the ellipse with the center circle of the circular target.

[0074] Furthermore, the specific testing methods are as follows:

[0075] Step S3.1: Binarize the grayscale image according to expression 2):

[0076] threshold=(minGray+aveGray) / rate 2),

[0077] Where threshold is the grayscale threshold for binarization; minGray is the minimum grayscale value of the grayscale image; aveGray is the average grayscale value of the grayscale image; rate is the ratio, and its value range is [2, 4].

[0078] Step S3.2: Find the four vertices of the minimum bounding rectangle of the ellipse based on the parameters of the ellipse, transform this minimum bounding rectangle into a square and solve for the affine transformation matrix, and perform an affine transformation on the binarized image based on the affine transformation matrix. After the affine transformation, the ellipse is transformed into a circle; the center of the circle after the affine transformation is (circleX, circleY), and the radius is radius.

[0079] Step S3.3: Since the center circle of the circular target is black, the pixel value inside the center circle is 0 after binarization. The area between the center circle and the coding ring is white, and the pixel value of this area should be 1 after binarization. Therefore, if the ellipse satisfies expression 3), it is a potential center circle.

[0080]

[0081] Wherein, innerPixels is the number of pixels inside the target circle, blackPixels is the number of pixels with a value of 0 inside the target circle, outerPixels is the number of pixels in the region between the target circle and the coding ring, whitePixels is the number of pixels with a value of 1 in the region between the target circle and the coding ring, and percent takes values ​​in the range of [0.8, 1].

[0082] Furthermore, the pixels (x, y) in innerPixels satisfy... The pixels (x, y) in outerPixels satisfy the following conditions:

[0083] Step S3.4, as follows Figure 3 As shown, with (circleX, circleY) as the center, A circular ring is formed with radius p. A pixel is taken every ang on the ring, resulting in a total of 360 / ang pixels p. i The subscript i represents the pixel number and its value range is [1, 360 / ang]; the value range of ang is [1, 5] and ang is divisible by N, and the unit is degrees. In this embodiment, ang is taken as 2.

[0084] Specifically, pixel p i The pixel coordinates are shown in expression 4):

[0085]

[0086] Obtain each pixel p in sequence i The corresponding pixel value g i Search for g i The number q corresponding to the first mutation site in the sequence (a list of 0s and 1s).

[0087] From g i Take from sequence Where k is an integer and its value ranges from 1 to 1. Total obtained Data sets; since N is taken as 30° in this embodiment, therefore: g i Take from sequence k is an integer and its value ranges from [0, 11];

[0088] Let count be the number of 0 values ​​in the k-th data set. k The encoded value (bit) of the k-th data group is shown in expression 5):

[0089]

[0090] exist In a data set: if any data set contains a negative bit value, then the ellipse is not the center circle of the circular target; if all data sets do not contain negative bits value, then the ellipse is the center circle of the circular target.

[0091] If the ellipse is the target center circle, then the value of this encoded ring is... The binary number corresponding to Code is cyclically shifted to obtain the minimum value minCode, which is used as the encoding value of the corresponding circular target.

[0092] Step S4: For the two identified circular targets, perform perspective transformation on the target area based on the common tangent of the ellipse to obtain the orthophoto image of the crack, calculate the actual distance between the centers of the two circular targets, and realize the monitoring of the crack.

[0093] Furthermore, step S4 is specifically as follows:

[0094] Step S4.1: Perform sub-pixel extraction on the ellipses EllipseA and EllipseB corresponding to the center circles of the two identified circular targets (the projection of the center circle onto the photo is an ellipse), and obtain the accurate ellipse equations as shown in expression 6):

[0095]

[0096] The preferred process for sub-pixel extraction and accurate ellipse equation acquisition is as follows:

[0097] Based on the identified ellipse (Ellipse A or Ellipse B), its general equation is obtained as: ax 2 +bxy+cy 2 +d x+e y+f=0;

[0098] Let the coordinates of a point on the ellipse be (ex, ey). Solve for the unit normal vector of the normal line corresponding to this point as (nX, nY). Take m pixels on each side of this point along the normal line direction. In this embodiment, m is taken as 5.

[0099] Let each pixel on both sides of the normal and the point (ex, ey) be denoted as point pj, where j = -m, -m+1, ..., m, and p0 represents point (ex, ey);

[0100] Calculate p for each point based on pixel value. j gray derivative g′ j , where g′ j =|g j -g j-1 |+|g j+1 -g j |,g j It is point p j Pixel values;

[0101] For the obtained g′ j Perform Gaussian curve fitting and find the mean extreme point of the corresponding Gaussian curve. The pixel coordinates corresponding to the mean extreme point are (gx, gy), where (gx, gy) are the sub-pixel coordinates of the point (ex, ey).

[0102]

[0103] Specifically, at least five points are selected on the ellipse to solve for their respective sub-pixel coordinates. The accurate ellipse equation is then obtained using the least squares method based on the obtained sub-pixel coordinate values.

[0104] Step S4.2: Obtain the equations of the common tangents between all common tangents of ellipses EllipseA and EllipseB, where the two external common tangents are l1 and l2, and the two internal common tangents are l3 and l4, as shown below. Figure 4 As shown;

[0105] Calculate the coordinates of the tangent points of the four common tangent lines with EllipseA: cutPt1, cutPt3, cutPt5, and cutPt7; and the coordinates of the tangent points of the four common tangent lines with EllipseB: cutPt2, cutPt4, cutPt6, and cutPt8.

[0106] Specifically, let the equation of the common tangent of the ellipse be:

[0107] y=λx+b 8),

[0108] Solve the simultaneous equations 6) and 8) to find λ and b corresponding to the four common tangents.

[0109] Step S4.3: Project the tangent points of the two external common tangents to the two ellipses EllipseA and EllipseB respectively onto the minimum bounding rectangles corresponding to the tangent points to construct the perspective transformation matrix M, and perform perspective transformation on the crack image according to the perspective transformation matrix M.

[0110] Specifically, the lines connecting the external common tangents l1 and l2 to the four tangent points cutPt1, cutPt2, cutPt4, and cutPt3 of the two ellipses form a matrix, which, when projected onto the photograph, becomes a polygon. Therefore, it is necessary to project these four tangent points onto their corresponding minimum bounding rectangles (referring to the minimum bounding rectangles corresponding to the four tangent points) to construct a perspective transformation matrix M. The crack image is then transformed using this perspective transformation matrix M. Since the aspect ratios of the rectangles corresponding to these four tangent points are not fixed, the crack image after perspective transformation is not an orthophoto, but rather scaled in a certain direction.

[0111] Step S4.4: Let the ellipses corresponding to EllipseA and EllipseB after perspective transformation be EllipseA′ and EllipseB′, respectively. Perform affine transformations on ellipses EllipseA′ and EllipseB′ respectively (the specific operation method is as in step S3.2) to obtain image A and image B.

[0112] Specifically, in image A, the center coordinates of the circle after the EllipseA′ transformation are (circleA_X1, circleA_Y1), the radius is radiusA, and the area of ​​the circle is areaA1. The center coordinates of the circle after the EllipseB′ transformation are (circleB_X1, circleB_Y1), and the area of ​​the circle is areaB1.

[0113] In image B, the center coordinates of the circle after EllipseA′ transformation are (circleA_X2, circleA_Y2), and the area of ​​the circle is areaA2. The center coordinates of the circle after EllipseB′ transformation are (circleB_X2, circleB_Y2), the radius is radiusB, and the area of ​​the circle is areaB2.

[0114] Step S4.5: Based on images A and B, calculate the actual distance between the centers of the two circular targets;

[0115] Specifically, the distance between the bullseyes of the two targets in image A is:

[0116]

[0117] The distance between the bullseye of the two targets in image B is:

[0118]

[0119] The actual distance between the centers of the two circular targets in

[0120] The effect of applying the technical solution of this invention is:

[0121] The method in this embodiment automatically calculates the width of cracks using paired circular targets, enabling precise monitoring of changes in cracks in buildings and structures. When taking photographs within 1 meter, the monitoring accuracy can reach the sub-millimeter level. It is simple to deploy on-site, low in cost, and ensures high monitoring accuracy, thus improving the automation level of crack monitoring.

[0122] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method of crack monitoring, characterized by, The method comprises the following steps: Step S1, a pair of circular targets are arranged on both sides of the crack, the circular targets comprising a coding ring and a target center circle inside the coding ring; Step S2, a crack image containing the crack and the pair of circular targets is obtained by a camera; the crack image is corrected and converted into a gray image; edge detection is performed on the gray image, and then ellipse searching is performed according to the detected contour; Step S3, in the ellipse identified in step S2, an ellipse that is a target center circle of the circular target is detected; Step S4, for the two identified circular targets, a perspective transformation is performed on the target area according to the common tangent of the ellipse, an orthographic image of the crack is obtained, the actual distance of the target centers of the two circular targets is calculated, and the monitoring of the crack is realized; The step S3 comprises the following steps: Step S3.1, the gray image is binarized according to expression 2): 2）； wherein, threshold is a binary threshold value of the grayscale image; minGray is a minimum grayscale value of the grayscale image; aveGray is an average grayscale value of the grayscale image; rate is a ratio, which takes a value in the range [2, 4]; Step S3.2, according to the parameters of the ellipse, four vertices of the minimum circumscribed rectangle of the ellipse are calculated, the minimum circumscribed rectangle is transformed into a square and an affine transformation matrix is solved, the binary image is subjected to affine transformation according to the affine transformation matrix, and the ellipse is transformed into a circle after the affine transformation; the center of the circle is circleX , circleY ) and the radius is radius . Step S3.3, if the ellipse satisfies expression 3), it is a potential target center circle: 3）； wherein, innerPixels is the number of pixels inside the bullseye circle, blackPixels is the number of pixels inside the bullseye circle having a pixel value of 0, outerPixels is the number of pixels in the area between the bullseye circle and the encoding ring, whitePixels is the number of pixels in the area between the bullseye circle and the encoding ring having a pixel value of 1, percent has a value in the range [0.8, 1]; Step S3.4, forming a circular ring with (R2, R3) as the center and R2 as the radius, and taking a pixel point every ang on the circular ring, to obtain 360 / ang pixel points circleX , circleY ) as the center and R2 as the radius , and taking a pixel point every ang on the circular ring, to obtain 360 / ang pixel points ; the subscript represents the number of the pixel point; ang takes a value in the range of [1, 5] and is an integer multiple of 10, in units of °; R2 is the radius of the inner circle of the encoding ring, and R3 is the radius of the outer circle of the encoding ring N ​ successively acquiring the pixel points corresponding pixel value , searching the number corresponding to the first mutation position of the sequence q ; From sequences are taken wherein k is an integer and takes the value range [0, ], obtaining data groups; N is the angle corresponding to the encoding bit in the encoding ring; The number of 0s in the first data group is denoted as k The number of 0s in the first data group is denoted as The number of 0s in the first data group is denoted as k The number of 0s in the first data group is denoted as As shown in expression 5): 5）； exist In a set of data groups, if the encoded value of any data group bit If the ellipse contains negative values, then it is not the center circle of a circular target; if the encoded values ​​of all data groups bit All values ​​are negative, and this ellipse is the center circle of a circular target.

2. The crack monitoring method according to claim 1, characterized by, The encoding ring totals 360°, with one encoding bit per degree, for a total of 360 encoding bits. N The encoding ring totals 360°, with one encoding bit per degree, for a total of 360 encoding bits. N The encoding ring totals 360°, with one encoding bit per degree, for The N is in the range [10, 60] and N is an integer division of 360, N is in degrees.

3. The crack monitoring method of claim 1, wherein In the step S2, if one contour contour is considered as an ellipse if it satisfies the expression 1): 1）； wherein, ptsNum is the contour contour is the number of points comprised in the contour polyDP is the contour contour is the number of vertices of the convex polygon fitted to the contour contourArea is the contour contour is the area of the contour rectArea is the contour contour is the area of the minimum circumscribed rectangle.

4. The crack monitoring method of claim 1, wherein After the ellipse is identified as the target circle, the value of the encoding ring is obtained as , and Code The minimum value is obtained by cyclically shifting the corresponding binary number minCode as the encoding value of the corresponding circular target.

5. The crack monitoring method of claim 1, wherein innerPixels the pixel point in the target circle satisfies ; ; outerPixels the pixel point in the target circle satisfies ; , R1 is the radius of the target circle.

6. The crack monitoring method of claim 1, wherein, The step S4 comprises the following steps: Step S4.1, sub-pixel extraction is performed on the ellipses EllipseA and EllipseB corresponding to the target center circles of the two identified circular targets, and accurate ellipse equations are obtained respectively; Step S4.2, common tangent equations of all common tangents between the ellipses EllipseA and EllipseB are obtained; Step S4.3, the two outer common tangents are respectively projected to the tangent points of the ellipses EllipseA and EllipseB to construct a perspective transformation matrix M, and perspective transformation is performed on the crack image according to the perspective transformation matrix M; Step S4.4, the perspective transformed ellipses corresponding to the ellipses Ellipse A and Ellipse B are respectively and , affine transformation is performed on the ellipses and , to obtain the images A and B; Step S4.5, the actual distance of the target centers of the two circular targets is calculated according to the images A and B.

7. The crack monitoring method of claim 6, wherein, In the step S4.1, according to the identified ellipse, a general ellipse equation thereof is obtained; at least 5 points on the ellipse are selected to solve their respective sub-pixel coordinates, and accurate ellipse equations are obtained through least square method according to the obtained sub-pixel coordinate values.

8. The crack monitoring method of claim 7, wherein, The solving method of the sub-pixel coordinates is as follows: Let the coordinates of a point on the ellipse be , the unit normal vector of the normal line corresponding to the point is , and take m pixel points on both sides of the normal line at the point; record each pixel point on both sides of the normal line and the point as point ; calculating the gray scale derivative of each point ; wherein ; is the pixel value of point ; The obtained Gaussian curve fitting is performed to obtain the extreme point mean of the corresponding Gaussian curve, and the pixel coordinates corresponding to the extreme point mean are wherein The sub-pixel coordinates of the point are 7）。 9. The crack monitoring method according to claim 6, characterized in that: In image A The transformed circle center coordinate is , the radius is radiusA, the area of the circle is areaA1, The transformed circle center coordinate is , the area of the circle is areaB1; the distance between the centers of the two circular targets in image A is: ， In image B The transformed circle center coordinate is , the area of the circle is areaA2, The transformed circle center coordinate is , the radius is radiusB, and the area of the circle is areaB2; the distance between the centers of the two circular targets in image B is: ， The actual distance of the two circular target centers ; wherein: , , R1 is the radius of the bullseye.

Images