A concrete crack detection method based on optical scanning technology

Abstract

The invention discloses a concrete crack detection method based on optical scanning technology, belonging to the technical field of image processing. The invention scans a concrete surface to obtain a concrete surface image, extracts contour pixel points thereon to obtain a contour map, and since there are multiple pixel points on the contour line on the contour map, the contour line is uneven in thickness, which affects crack recognition, and therefore, refines the contour line to obtain three refined contour maps with clear lines, and reduces the line width to one pixel point so as to more accurately represent the actual form of the crack, and then performs shearing processing on the three refined contour maps to screen out irregular image blocks, and reconstructs each position point according to each irregular image block to obtain cracks, thereby improving the accuracy of crack detection.

Description

Technical Field

[0001] The invention relates to the technical field of image processing, and in particular to a concrete crack detection method based on optical scanning technology. Background Art

[0002] As a widely used building material, the durability and stability of concrete are crucial to the safety of engineering structures. However, due to various reasons, such as material aging, external forces, environmental erosion, etc., cracks may appear in concrete structures during use. These cracks not only affect the appearance of the structure, but may also become channels for the penetration of moisture and harmful chemicals, further accelerating the damage of the structure. Therefore, timely detection and evaluation of concrete cracks are of great significance to ensure the safety of the structure and extend its service life.

[0003] Traditional crack detection methods often rely on manual visual inspection, which is not only inefficient, but also has accuracy affected by subjective factors. With the development of technology, automated crack detection methods based on optical scanning technology have gradually attracted attention. These methods use high-precision optical equipment to scan the concrete surface and analyze crack characteristics through image processing technology to achieve fast and accurate crack detection. Sobel, Canny and other operators are commonly used to detect edges in images, and cracks are identified by analyzing edge information, but they are easily affected by noise and have the problem of low crack detection accuracy. Summary of the invention

[0004] In view of the above-mentioned deficiencies in the prior art, the present invention provides a concrete crack detection method based on optical scanning technology to solve the problem of low crack detection accuracy in the prior art.

[0005] In order to achieve the above-mentioned purpose of the invention, the technical solution adopted by the present invention is: a concrete crack detection method based on optical scanning technology, comprising the following steps:

[0006] S1, scanning the concrete surface with an optical acquisition device to obtain a concrete surface image;

[0007] S2, extracting contour pixels from the concrete surface image to obtain a contour map;

[0008] S3, thinning the contour lines on the contour map to obtain a first thinned contour map, a second thinned contour map and a third thinned contour map, wherein the first thinned contour map is composed of the maximum position points of each local area, the second thinned contour map is composed of the minimum position points of the local area, and the third thinned contour map is composed of the center position points of the local area;

[0009] S4, performing shearing processing on the first refined contour map, the second refined contour map, and the third refined contour map, constructing a morphological vector for the sheared image blocks, and screening out irregular image blocks;

[0010] S5. Reconstruct each position point according to each irregular image block to obtain a crack.

[0011] Furthermore, the S2 comprises the following sub-steps:

[0012] S21, taking each pixel point on the concrete surface image as the center, calculating the distribution coefficient of the pixel values ​​of four pixel points within the neighborhood of each central pixel point to obtain a first distribution coefficient;

[0013] S22, calculating the distribution coefficients of the pixel values ​​of the other four pixels within the neighborhood of each central pixel to obtain a second distribution coefficient;

[0014] S23, when the first distribution coefficient and the second distribution coefficient are both greater than the coefficient threshold, the central pixel point is a contour pixel point;

[0015] S24, all the contour pixels are combined to form a contour map.

[0016] Furthermore, the formulas for calculating the distribution coefficient in S21 and S22 are both:

[0017] ,

[0018] Wherein, θ is the first distribution coefficient or the second distribution coefficient, P1 is the pixel value of the first pixel point in the neighborhood range, P2 is the pixel value of the second pixel point in the neighborhood range, P3 is the pixel value of the third pixel point in the neighborhood range, and P4 is the pixel value of the fourth pixel point in the neighborhood range. is the pixel mean of 4 pixels, and || is the absolute value operation.

[0019] Furthermore, S3 includes the following sub-steps:

[0020] S31. Settings The search window of size slides on each contour of the contour map. During each slide, the forward step length is 1 pixel, and K is a positive integer.

[0021] S32. After each slide, calculate the pixel distance for each contour pixel point covered by the search window: d i = (x i ) 2 + (y i ) 2 , d i is the pixel distance of the i-th contour pixel, x i is the horizontal coordinate of the i-th contour pixel, yi is the ordinate of the i-th contour pixel, i is a positive integer;

[0022] S33, after each sliding, the maximum position point, the minimum position point and the center position point of the contour pixel points covered by the search window are taken, wherein the maximum position point is max{d i} corresponding pixel point, the minimum position point is min{ d i} The center point is the geometric center point of each contour pixel point covered by the search window, max is the maximum value, and min is the minimum value;

[0023] S34, constructing a first refined contour map with each maximum position point;

[0024] S35, forming a second refined contour map with each minimum position point;

[0025] S36, forming a third refined contour map from each central position point.

[0026] Furthermore, the S4 comprises the following sub-steps:

[0027] S41, cutting the first refined contour map, the second refined contour map and the third refined contour map in the same manner to obtain the first refined contour map corresponding to image blocks, the second refined contour map corresponds to image blocks, the third refined contour map corresponds to image blocks, N is a positive integer;

[0028] S42, constructing a Hankel matrix for the pixel distances of each position point of each image block;

[0029] S43, calculating the morphology value of each row according to the Hankel matrix, and forming a morphology vector with the morphology values ​​of each row;

[0030] S44, calculating the irregularity of the image block according to the morphological vector;

[0031] S45: When the irregularity of the image block is greater than the irregularity threshold, the image block is an irregular image block.

[0032] Furthermore, the formula for calculating the morphology value of each row in S43 is:

[0033] ,

[0034] Among them, ε k is the shape value of the kth row, e is a natural constant, x k,j is the jth element in the kth row of the Hankel matrix, M is the number of elements in each row, and k and j are positive integers.

[0035] Furthermore, the formula for calculating the irregularity of the image block in S44 is:

[0036] ,

[0037] Among them, γ is the irregularity of the image block, ε k is the shape value of the kth row, ε th is the morphological threshold, μ is the variance of the morphological vector, k is a positive integer, and L is the number of rows of the Hankel matrix.

[0038] Furthermore, the S5 comprises the following sub-steps:

[0039] S51, taking the intersection of the irregular image blocks corresponding to the first refined contour map, the irregular image blocks corresponding to the second refined contour map, and the irregular image blocks corresponding to the third refined contour map to obtain image blocks in the same area, and marking the image blocks in the same area as crack blocks;

[0040] S52, splicing the crack blocks on the first refined contour map to obtain a first spliced ​​contour;

[0041] S53, splicing the crack blocks on the second refined contour map to obtain a second spliced ​​contour;

[0042] S54, splicing the crack blocks on the third refined contour map to obtain a third spliced ​​contour;

[0043] S55. Reconstruct each position point according to the first splicing contour, the second splicing contour and the third splicing contour to obtain a crack.

[0044] Furthermore, the S55 comprises the following sub-steps:

[0045] S551, reconstructing a new position point according to the maximum position point on the first spliced ​​contour, the minimum position point on the second spliced ​​contour, and the center position point on the third spliced ​​contour;

[0046] S552, each new position point constitutes a crack.

[0047] Furthermore, the reconstruction formula in S551 is:

[0048] ,

[0049] Among them, (x, y) is the new position point, x is the horizontal coordinate of the new position point, y is the vertical coordinate of the new position point, (x max ,y max ) is the maximum position point, x max is the horizontal coordinate of the maximum position point, y max is the ordinate of the maximum position point, (xmin ,y min ) is the minimum position point, x min is the horizontal coordinate of the minimum position point, y min is the ordinate of the minimum position point, (x o ,y o ) is the center point, x o is the horizontal coordinate of the center point, y o is the ordinate of the center point, α1 is the first weighting coefficient, α2 is the second weighting coefficient, and α3 is the third weighting coefficient.

[0050] The beneficial effects of the present invention are as follows: the present invention scans the concrete surface to obtain a concrete surface image, extracts contour pixel points therefrom to obtain a contour map, and since there are multiple pixel points on the contour line on the contour map, the contour line is uneven in thickness, which affects crack recognition, and therefore, refines the contour line to obtain three refined contour maps with clear lines, and reduces the line width to one pixel point so as to more accurately represent the actual shape of the crack, and then performs shearing processing on the three refined contour maps to screen out irregular image blocks, and reconstructs each position point according to each irregular image block to obtain cracks, thereby improving the accuracy of crack detection.

[0051] When the contour lines on the contour map are refined, the present invention extracts three refined contour maps, so that when the irregular image blocks are subsequently screened, there are three processing objects, which greatly reduces the influence of noise and improves the accuracy of crack screening. At the same time, each position point is reconstructed according to each irregular image block on the three refined contour maps to accurately depict the shape of the crack. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 is a flow chart of a concrete crack detection method based on optical scanning technology;

[0053] Figure 2 is a schematic diagram of the concrete surface image;

[0054] Figure 3 It is a schematic diagram of the contour map;

[0055] Figure 4 is a schematic diagram of a first refined contour map;

[0056] Figure 5 is a schematic diagram of the second refined contour map;

[0057] Figure 6 It is a schematic diagram of the third refined contour map. DETAILED DESCRIPTION

[0058] The specific implementation modes of the present invention are described below so that those skilled in the art can understand the present invention. However, it should be clear that the present invention is not limited to the scope of the specific implementation modes. For those of ordinary skill in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the attached claims, these changes are obvious, and all inventions and creations utilizing the concept of the present invention are protected.

[0059] like Figure 1 As shown, a concrete crack detection method based on optical scanning technology includes the following steps:

[0060] S1. Scan the concrete surface with optical equipment to obtain the concrete surface image, such as Figure 2 As shown;

[0061] S2, extracting contour pixels from the concrete surface image to obtain a contour map;

[0062] S3, thinning the contour lines on the contour map to obtain a first thinned contour map, a second thinned contour map and a third thinned contour map, wherein the first thinned contour map is composed of the maximum position points of each local area, the second thinned contour map is composed of the minimum position points of the local area, and the third thinned contour map is composed of the center position points of the local area;

[0063] S4, performing shearing processing on the first refined contour map, the second refined contour map, and the third refined contour map, constructing a morphological vector for the sheared image blocks, and screening out irregular image blocks;

[0064] S5. Reconstruct each position point according to each irregular image block to obtain a crack.

[0065] The S2 comprises the following sub-steps:

[0066] S21, taking each pixel point on the concrete surface image as the center, calculating the distribution coefficient of the pixel values ​​of four pixel points within the neighborhood of each central pixel point to obtain a first distribution coefficient;

[0067] S22, calculating the distribution coefficients of the pixel values ​​of the other four pixels within the neighborhood of each central pixel to obtain a second distribution coefficient;

[0068] S23, when the first distribution coefficient and the second distribution coefficient are both greater than the coefficient threshold, the central pixel point is a contour pixel point;

[0069] S24, all the contour pixels are used to form a contour map, such as Figure 3 shown.

[0070] A central pixel point has 8 neighboring pixel points. The present invention divides the 8 neighboring pixel points into two parts, and calculates the first distribution coefficient and the second distribution coefficient respectively. When the first distribution coefficient and the second distribution coefficient are both greater than the coefficient threshold, the central pixel point is a contour pixel point, which improves the accuracy of screening contour pixels and avoids the influence of noise points on the extraction of contour pixels.

[0071] The formulas for calculating the distribution coefficient in S21 and S22 are both:

[0072] ,

[0073] Wherein, θ is the first distribution coefficient or the second distribution coefficient, P1 is the pixel value of the first pixel point in the neighborhood range, P2 is the pixel value of the second pixel point in the neighborhood range, P3 is the pixel value of the third pixel point in the neighborhood range, and P4 is the pixel value of the fourth pixel point in the neighborhood range. is the pixel mean of 4 pixels, and || is the absolute value operation.

[0074] The present invention calculates the absolute value of the difference between the pixel value of each pixel point in the neighborhood range and the pixel mean, reflecting the distribution of the pixel values ​​of the four pixel points. When the distribution coefficient is larger, the contour feature is more obvious.

[0075] In the present invention, the coefficient threshold is a threshold set for the distribution coefficient, and its specific value can be specifically set based on experiments or experience.

[0076] In this embodiment, the method of extracting the contour map in step S2 also includes: hollowing out the interior method, boundary tracking method, region growing method, etc.

[0077] The S3 comprises the following sub-steps:

[0078] S31. Settings The search window of size slides on each contour of the contour map. During each slide, the forward step length is 1 pixel, and K is a positive integer.

[0079] S32. After each slide, calculate the pixel distance for each contour pixel point covered by the search window: d i = (x i ) 2 + (y i ) 2 , d i is the pixel distance of the i-th contour pixel, x i is the horizontal coordinate of the i-th contour pixel, y i is the ordinate of the i-th contour pixel, i is a positive integer;

[0080] S33, after each sliding, the maximum position point, the minimum position point and the center position point of the contour pixel points covered by the search window are taken, wherein the maximum position point is max{d i} corresponding pixel point, the minimum position point is min{ d i} The center point is the geometric center point of each contour pixel point covered by the search window, max is the maximum value, and min is the minimum value;

[0081] S34, each maximum position point constitutes a first refined contour map, such as Figure 4 As shown;

[0082] S35, each minimum position point constitutes a second refined contour map, such as Figure 5 As shown;

[0083] S36, each center position point forms a third refined contour map, such as Figure 6 shown.

[0084] The present invention sets A search window of different sizes is slid on each contour. During each slide, the maximum position point, minimum position point and center position point are selected according to the multiple contour pixel points covered by the search window. The maximum position point and the minimum position point represent the outermost edge of the contour, and the center position point represents the center of the contour. Therefore, the maximum position point, the minimum position point and the center position point are used to construct a refined contour map respectively, so that the uneven contour is transformed into a contour with a width of 1 pixel.

[0085] In this embodiment, the value range of K is 2, 3, 4, and 5.

[0086] The S4 comprises the following sub-steps:

[0087] S41, cutting the first refined contour map, the second refined contour map and the third refined contour map in the same manner to obtain the first refined contour map corresponding to image blocks, the second refined contour map corresponds to image blocks, the third refined contour map corresponds to image blocks, N is a positive integer;

[0088] S42, constructing a Hankel matrix for the pixel distances of each position point of each image block;

[0089] S43, calculating the morphology value of each row according to the Hankel matrix, and forming a morphology vector with the morphology values ​​of each row;

[0090] S44, calculating the irregularity of the image block according to the morphological vector;

[0091] S45: When the irregularity of the image block is greater than the irregularity threshold, the image block is an irregular image block.

[0092] In the present invention, the pixel distances are calculated using the method disclosed in S32.

[0093] The present invention cuts the three refined contour images in the same way to obtain image blocks, that is, each refined contour map is sheared N-1 times horizontally, and then the refined contour map is sheared N-1 times vertically, which helps to analyze different parts of the crack separately, thereby improving the accuracy of crack recognition. The present invention constructs a Hankel matrix for the pixel distance of each position point of the image block, and uses the gradually progressive characteristics of each row of elements in the Hankel matrix to calculate the morphological value of each row. The morphological value reflects the morphological characteristics of a section of position points, and then calculates the irregularity of the image block according to the morphological vector, and screens out the image block with an irregularity greater than a threshold. The contour lines contained in the image block are extremely irregular. By calculating the irregularity of the image block and identifying the image block with an irregularity greater than a threshold as an irregular image block, cracks and non-crack areas can be effectively distinguished.

[0094] In this embodiment, N can be set to a positive integer greater than or equal to 10.

[0095] The irregularity threshold is a threshold set for the irregularity, and the specific value is set based on experiments or experience.

[0096] The formula for calculating the morphological value of each row in S43 is:

[0097] ,

[0098] Among them, ε k is the shape value of the kth row, e is a natural constant, x k,j is the jth element in the kth row of the Hankel matrix, M is the number of elements in each row, and k and j are positive integers.

[0099] The present invention adopts the element of each row minus the mean value of each row, and then adopts the exponential function to enhance the difference, so as to reflect the fluctuation of each element.

[0100] In this embodiment, the method of calculating the morphological value of each row includes: variance and standard deviation.

[0101] The formula for calculating the irregularity of the image block in S44 is:

[0102] ,

[0103] Among them, γ is the irregularity of the image block, ε k is the shape value of the kth row, ε this the morphological threshold, μ is the variance of the morphological vector, k is a positive integer, and L is the number of rows of the Hankel matrix.

[0104] The present invention sets a morphology threshold value for the morphology value, so as to measure the morphology of each row of elements. When the morphology value is larger, the irregularity is higher. At the same time, the variance is used to measure the irregularity of the entire morphology vector, so as to further enhance the irregularity and improve the accuracy of screening irregular image blocks.

[0105] In the present invention, the morphology threshold is a threshold value set for the morphology value, and the specific value is specifically set based on experiments or experience.

[0106] The S5 comprises the following sub-steps:

[0107] S51, taking the intersection of the irregular image blocks corresponding to the first refined contour map, the irregular image blocks corresponding to the second refined contour map, and the irregular image blocks corresponding to the third refined contour map to obtain image blocks in the same area, and marking the image blocks in the same area as crack blocks;

[0108] S52, splicing the crack blocks on the first refined contour map to obtain a first spliced ​​contour;

[0109] S53, splicing the crack blocks on the second refined contour map to obtain a second spliced ​​contour;

[0110] S54, splicing the crack blocks on the third refined contour map to obtain a third spliced ​​contour;

[0111] S55. Reconstruct each position point according to the first splicing contour, the second splicing contour and the third splicing contour to obtain a crack.

[0112] Since the three refined contour maps are derived from the same contour map, and the three refined contour maps use the same cutting method, the irregular image blocks on the three refined contour maps are intersected, that is, image blocks with cracks in the same area are found, and the crack blocks on the three refined contour maps are spliced ​​according to the original positions of the crack blocks in the refined contour map to obtain three spliced ​​contours. The positions on the three spliced ​​contours are reconstructed to obtain cracks.

[0113] The S55 comprises the following sub-steps:

[0114] S551, reconstructing a new position point according to the maximum position point on the first spliced ​​contour, the minimum position point on the second spliced ​​contour, and the center position point on the third spliced ​​contour;

[0115] S552, each new position point constitutes a crack.

[0116] The reconstruction formula in S551 is:

[0117] ,

[0118] Among them, (x, y) is the new position point, x is the horizontal coordinate of the new position point, y is the vertical coordinate of the new position point, (x max ,y max ) is the maximum position point, x max is the horizontal coordinate of the maximum position point, y max is the ordinate of the maximum position point, (x min ,y min ) is the minimum position point, x min is the horizontal coordinate of the minimum position point, y min is the ordinate of the minimum position point, (x o ,y o ) is the center point, x o is the horizontal coordinate of the center point, y o is the ordinate of the center point, α1 is the first weighting coefficient, α2 is the second weighting coefficient, and α3 is the third weighting coefficient.

[0119] When the present invention is reconstructing, when reconstructing each new position point, the maximum position point, the minimum position point and the center position point selected are the three position points extracted by the search window after each sliding in step S33, that is, it is ensured that during reconstruction, the maximum position point, the minimum position point and the center position point are points within the same range.

[0120] In this embodiment, the contour on the third stitching contour can also be used as a reference to select a center position point from the third stitching contour, calculate the position distance between each center position point and the maximum position point on the first stitching contour and the minimum position point on the second stitching contour, filter out the maximum position point and the minimum position point with the closest distance, and reconstruct the maximum position point, the minimum position point, and the corresponding center position point into a new position point.

[0121] In this embodiment, α1+α2+α3=1, α1=0.25, α2=0.25, α3=0.5.

[0122] The present invention scans the concrete surface to obtain a concrete surface image, extracts contour pixel points thereon to obtain a contour map, and since there are multiple pixel points on the contour line of the contour map, the contour line is uneven in thickness, which affects crack recognition. Therefore, it is refined to obtain three refined contour maps with clear lines, and the line width is reduced to one pixel point to more accurately represent the actual shape of the crack. The three refined contour maps are then sheared to screen out irregular image blocks, and each position point is reconstructed according to each irregular image block to obtain cracks, thereby improving the accuracy of crack detection.

[0123] When the contour lines on the contour map are refined, the present invention extracts three refined contour maps, so that when the irregular image blocks are subsequently screened, there are three processing objects, which greatly reduces the influence of noise and improves the accuracy of crack screening. At the same time, each position point is reconstructed according to each irregular image block on the three refined contour maps to accurately depict the shape of the crack.

[0124] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A concrete crack detection method based on optical scanning technology, characterized in that: The following steps are involved: S1, scanning the concrete surface with an optical acquisition device to obtain a concrete surface image; S2, extracting contour pixels from the concrete surface image to obtain a contour map; S3, thinning the contour lines on the contour map to obtain a first thinned contour map, a second thinned contour map and a third thinned contour map, wherein the first thinned contour map is composed of the maximum position points of each local area, the second thinned contour map is composed of the minimum position points of the local area, and the third thinned contour map is composed of the center position points of the local area; S4, performing shearing processing on the first refined contour map, the second refined contour map, and the third refined contour map, constructing a morphological vector for the sheared image blocks, and screening out irregular image blocks; S5. Reconstruct each position point according to each irregular image block to obtain a crack; The S4 comprises the following sub-steps: S41, cutting the first refined contour map, the second refined contour map and the third refined contour map in the same manner to obtain the first refined contour map corresponding to image blocks, the second refined contour map corresponds to image blocks, the third refined contour map corresponds to image blocks, N is a positive integer; S42, constructing a Hankel matrix for the pixel distances of each position point of each image block; S43, calculating the morphology value of each row according to the Hankel matrix, and forming a morphology vector with the morphology values ​​of each row; S44, calculating the irregularity of the image block according to the morphological vector; S45: when the irregularity of the image block is greater than the irregularity threshold, the image block is an irregular image block; The formula for calculating the morphological value of each row in S43 is: ， Among them, ε k is the shape value of the kth row, e is a natural constant, x k,j is the jth element in the kth row of the Hankel matrix, M is the number of elements in each row, and k and j are positive integers.

2. The concrete crack detection method based on optical scanning technology according to claim 1 is characterized in that: The S2 comprises the following sub-steps: S21, taking each pixel point on the concrete surface image as the center, calculating the distribution coefficient of the pixel values ​​of four pixel points within the neighborhood of each central pixel point to obtain a first distribution coefficient; S22, calculating the distribution coefficients of the pixel values ​​of the other four pixels within the neighborhood of each central pixel to obtain a second distribution coefficient; S23, when the first distribution coefficient and the second distribution coefficient are both greater than the coefficient threshold, the central pixel point is a contour pixel point; S24, all the contour pixels are combined to form a contour map.

3. The concrete crack detection method based on optical scanning technology according to claim 2 is characterized in that: The formulas for calculating the distribution coefficient in S21 and S22 are both: ， Wherein, θ is the first distribution coefficient or the second distribution coefficient, P1 is the pixel value of the first pixel point in the neighborhood range, P2 is the pixel value of the second pixel point in the neighborhood range, P3 is the pixel value of the third pixel point in the neighborhood range, and P4 is the pixel value of the fourth pixel point in the neighborhood range. is the pixel mean of 4 pixels, and || is the absolute value operation.

4. The concrete crack detection method based on optical scanning technology according to claim 1 is characterized in that: The S3 comprises the following sub-steps: S31. Settings The search window of size slides on each contour of the contour map. During each slide, the forward step length is 1 pixel, and K is a positive integer. S32. After each slide, calculate the pixel distance for each contour pixel point covered by the search window: d i = (x i ) 2 +(y i ) 2 , d i is the pixel distance of the i-th contour pixel, x i is the horizontal coordinate of the i-th contour pixel, y i is the ordinate of the i-th contour pixel, i is a positive integer; S33, after each sliding, the maximum position point, the minimum position point and the center position point of the contour pixel points covered by the search window are taken, wherein the maximum position point is max{d i } corresponding pixel point, the minimum position point is min{ d i } The center point is the geometric center point of each contour pixel point covered by the search window, max is the maximum value, and min is the minimum value; S34, constructing a first refined contour map with each maximum position point; S35, forming a second refined contour map with each minimum position point; S36, forming a third refined contour map from each central position point.

5. The concrete crack detection method based on optical scanning technology according to claim 1 is characterized in that: The formula for calculating the irregularity of the image block in S44 is: ， Among them, γ is the irregularity of the image block, ε k is the shape value of the kth row, ε th is the morphological threshold, μ is the variance of the morphological vector, k is a positive integer, and L is the number of rows of the Hankel matrix.

6. The concrete crack detection method based on optical scanning technology according to claim 1 is characterized in that: The S5 comprises the following sub-steps: S51, taking the intersection of the irregular image blocks corresponding to the first refined contour map, the irregular image blocks corresponding to the second refined contour map, and the irregular image blocks corresponding to the third refined contour map to obtain image blocks in the same area, and marking the image blocks in the same area as crack blocks; S52, splicing the crack blocks on the first refined contour map to obtain a first spliced ​​contour; S53, splicing the crack blocks on the second refined contour map to obtain a second spliced ​​contour; S54, splicing the crack blocks on the third refined contour map to obtain a third spliced ​​contour; S55. Reconstruct each position point according to the first splicing contour, the second splicing contour and the third splicing contour to obtain a crack.

7. The concrete crack detection method based on optical scanning technology according to claim 6 is characterized in that: The S55 comprises the following sub-steps: S551, reconstructing a new position point according to the maximum position point on the first spliced ​​contour, the minimum position point on the second spliced ​​contour, and the center position point on the third spliced ​​contour; S552, each new position point constitutes a crack.

8. The concrete crack detection method based on optical scanning technology according to claim 7 is characterized in that: The reconstruction formula in S551 is: ， Among them, (x, y) is the new position point, x is the horizontal coordinate of the new position point, y is the vertical coordinate of the new position point, (x max ,y max ) is the maximum position point, x max is the horizontal coordinate of the maximum position point, y max is the ordinate of the maximum position point, (x min ,y min ) is the minimum position point, x min is the horizontal coordinate of the minimum position point, y min is the ordinate of the minimum position point, (x o ,y o ) is the center point, x o is the horizontal coordinate of the center point, y o is the ordinate of the center point, α1 is the first weighting coefficient, α2 is the second weighting coefficient, and α3 is the third weighting coefficient.

Images