Underground structure outer wall defect visual detection method and system

Abstract

The invention relates to the technical field of image detection, in particular to an underground structure outer wall defect visual detection method and system.The method comprises the steps that according to the difference between the length and the width of a minimum external rectangle of each connected domain and the included angle between the left side edge of the minimum external rectangle and the horizontal line where the bottom edge of a basement outer wall is located, the minimum external rectangle is determined; determining a morphological anomaly index of each connected domain; based on the distance from each pixel point on the framework in each connected domain to the left side edge and the right side edge of the minimum enclosing rectangle of the corresponding connected domain and the bending degree of each pixel point on the framework of each connected domain, the crack confidence coefficient of each connected domain is determined, and the basement exterior wall surface is detected. According to the method, the texture features of the cracks and the distribution features of the cracks and surrounding water spots are comprehensively analyzed, so that the accuracy of basement exterior wall surface crack detection is improved.

Description

Technical Field

[0001] The present application relates to the field of image detection technology, and in particular to a method and system for visually detecting defects in the exterior walls of underground structures. Background Art

[0002] As the development of underground space gradually increases, the number of underground structures put into construction and operation stages is increasing. However, due to the influence of natural environment, human activities, building materials and other factors, underground structures will inevitably have defects such as cracks, voids, and steel corrosion. These defects change the mechanical properties of underground structures to varying degrees, which may lead to safety accidents and damage people’s property safety. Therefore, efficient and timely detection of underground structure defects during the construction and operation periods has become a key task that needs to be carried out.

[0003] Underground structure defect detection is of great significance for ensuring the safety of underground structures, extending the service life of underground structures, protecting property safety, and improving the quality of buildings. It is an indispensable part of underground structure management and maintenance. In recent years, with the development of high-resolution imaging technology, the potential of image-based visual detection methods in underground structure defect detection has gradually increased. High-resolution imaging technology can accurately capture the surface information of underground structures, enabling inspectors to effectively identify surface defects on the exterior walls of underground structures. However, in practical applications, the accuracy of underground structure exterior wall defect detection is reduced due to the complex and uneven surface of the exterior walls. Summary of the invention

[0004] In order to solve the above technical problems, the purpose of this application is to provide a method and system for visual detection of defects in the exterior wall of an underground structure. The technical solutions adopted are as follows:

[0005] In a first aspect, an embodiment of the present application provides a method for visually detecting defects in an outer wall of an underground structure, the method comprising the following steps:

[0006] Obtain the grayscale image and binary image of the outer wall surface of each basement, use the result of bitwise multiplication of the grayscale image and the binary image as the target image, extract all connected domains in the target image, and obtain the minimum circumscribed rectangle of each connected domain;

[0007] According to the aspect ratio of the minimum circumscribed rectangle of each connected domain and the angle between the left edge of the minimum circumscribed rectangle and the horizontal line where the bottom edge of the basement outer wall is located, the morphological anomaly index of each connected domain is determined;

[0008] A skeleton extraction algorithm is used to obtain the skeleton in each connected domain. Based on the distance from each pixel point on the skeleton in each connected domain to the left and right edges of the minimum circumscribed rectangle of the corresponding connected domain, the symmetry difference of each pixel point on the skeleton in each connected domain is determined. Combined with the degree of curvature at each pixel point on the skeleton of each connected domain, the irregularity index of each connected domain is determined. Based on the irregularity index and the morphological anomaly index, the crack confidence of each connected domain is determined.

[0009] Based on the average distribution of crack confidences of all connected domains in the target image, the crack defect degree of each basement exterior wall surface is determined; based on the distribution of crack defect degrees of all basement exterior wall surfaces, the crack assessment threshold is determined, and the basement exterior wall surface to be detected is detected.

[0010] Preferably, the expression of the morphological abnormality index of each connected domain is: In the formula, CMA i represents the morphological abnormality index of the connected domain i; h i ,d i Respectively represent the length and width of the smallest external rectangle of the connected domain i; θ i Represents the angle between the left edge of the minimum circumscribed rectangle of the connected domain i and the horizontal line where the bottom edge of the basement outer wall is located; tan() represents the tangent function.

[0011] Preferably, the skeleton extraction algorithm is the Zhang Suen thinning algorithm.

[0012] Preferably, the expression of the symmetry difference degree of each pixel point on the skeleton in each connected domain is: i,m =|l i,m -r i,m |; In the formula, D i,m represents the symmetric difference of pixel point m on the skeleton within the connected domain i; l i,m 、r i,m They respectively represent the distances from the pixel point m on the skeleton within the connected domain i to the left and right edges of the minimum outer rectangle of the connected domain i.

[0013] Preferably, the expression of the irregularity index of each connected domain is: Where, CU i represents the irregular index of the connected domain i; q i,m represents the curvature of pixel m on the skeleton within the connected domain i; M i represents the number of all pixels on the skeleton within the connected domain i; β represents a constant preset to be greater than 0.

[0014] Preferably, the crack confidence of each connected domain is the product of the irregularity index and the morphological anomaly index of each connected domain.

[0015] Preferably, the crack defect degree of each basement outer wall surface is the average of the crack confidences of all connected domains in the target image of each basement outer wall surface.

[0016] Preferably, the crack assessment threshold is the average of the crack defect degrees of a preset number of basement exterior wall surfaces.

[0017] Preferably, the detection of the surface of the outer wall of the basement to be detected includes:

[0018] A grayscale image and a binary image of the exterior wall surface of the basement to be inspected are obtained, and the crack defect degree of the exterior wall surface of the basement to be inspected is obtained according to the calculation method of the crack defect degree of each outdoor wall surface. If the crack defect degree of the exterior wall surface of the basement to be inspected is greater than the crack assessment threshold, then there is a crack defect on the exterior wall surface of the basement to be inspected; otherwise, there is no crack defect on the exterior wall surface of the basement to be inspected.

[0019] In the second aspect, an embodiment of the present application also provides a visual detection system for underground structure exterior wall defects, including a memory, a processor, and a computer program stored in the memory and running on the processor, and when the processor executes the computer program, it implements the steps of any one of the above-mentioned methods for visual detection of underground structure exterior wall defects.

[0020] This application has at least the following beneficial effects:

[0021] The present application constructs a connected domain morphological anomaly index by analyzing the growth morphological characteristics of the crack area when cracks occur on the surface of the basement outer wall. The beneficial effect is that the difference between the morphology of the cracks on the basement outer wall and the connected domain formed by the normal wall surface is taken into account. Based on the morphological anomaly index, the morphological characteristics of the crack area on the basement outer wall surface can be better reflected, which is helpful to distinguish the connected domain where the cracks are located from the connected domain where the wall grooves are located, thereby reducing misjudgment. The present application constructs a crack confidence by analyzing the water seepage phenomenon of the cracks on the surface of the basement outer wall and the difference between the cracks and the wall grooves on the surface of the basement outer wall and combining the morphological anomaly index. The beneficial effect is that the crack confidence of the connected domain takes into account the left-right symmetrical structure of the water stains with the cracks as the center when water seeps into the crack area on the surface of the basement outer wall, as well as the irregular phenomenon of the cracks on the basement outer wall due to the different stresses of various parts of the wall, which helps to reduce the false alarm rate and ensure the accuracy of the detection results. The present application improves the accuracy of the detection of cracks on the surface of the basement outer wall by comprehensively analyzing the texture characteristics of the cracks and the distribution characteristics of the cracks and the surrounding water stains. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present application or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0023] Figure 1 A flowchart of a method for visually detecting defects in an outer wall of an underground structure provided in one embodiment of the present application;

[0024] Figure 2 A schematic diagram of an underground structure exterior wall image acquisition and processing process provided in one embodiment of the present application;

[0025] Figure 3 A schematic diagram of a crack confidence extraction process provided by an embodiment of the present application. DETAILED DESCRIPTION

[0026] In order to further explain the technical means and effects adopted by the present application to achieve the predetermined invention purpose, the following is a detailed description of the method and system for visual detection of defects in the exterior wall of an underground structure proposed in the present application, its specific implementation method, structure, features and effects, in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0027] Unless defined otherwise, 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 application belongs.

[0028] The specific scheme of a method and system for visual detection of defects in the exterior wall of an underground structure provided by the present application is described in detail below with reference to the accompanying drawings.

[0029] See also Figure 1 , which shows a flowchart of a method for visually detecting defects in an underground structure exterior wall provided by an embodiment of the present application, the method comprising the following steps:

[0030] Step S1: Obtain the grayscale image and binary image of each basement exterior wall surface, use the result of bitwise AND of the grayscale image and the binary image as the target image, extract all connected domains in the target image, and obtain the minimum circumscribed rectangle of each connected domain.

[0031] (1) Under good light conditions, a CCD camera is used to capture the basement exterior wall, and a preset number of basement exterior wall surface images are captured. In order to facilitate the detection of crack defects on the basement exterior wall surface and reduce subsequent calculation costs, the captured basement exterior wall surface images are grayed to obtain a basement exterior wall surface grayscale image. The value of the preset number is artificially set. In this embodiment, the value of the preset number is 100. The implementer can also set it by himself according to the specific situation. This embodiment does not impose any special restrictions.

[0032] (2) Furthermore, in order to reduce the significant impact of random noise in the shooting and collection working environment on the subsequent basement exterior wall surface crack defect detection, a filtering algorithm is used to perform noise reduction on the grayscale image of the basement exterior wall surface.

[0033] It should be noted that there are many commonly used filtering algorithms. In this embodiment, a Gaussian filtering algorithm is used to perform noise reduction on the grayscale image of the basement exterior wall surface. The implementer may also use other filtering methods such as median filtering or mean filtering algorithms. This embodiment does not impose any special restrictions on the selection of filtering algorithms.

[0034] Among them, the Gaussian filtering algorithm is a well-known technology in the field of noise reduction, and its specific noise reduction principle will not be described in detail.

[0035] (3) Since the crack itself is a concave structure, light entering it will be partially absorbed, and due to the water seepage in the crack, the grayscale value of the crack area in the basement exterior wall surface image is lower than that of the normal wall surface. In addition, the larger the width of the basement exterior wall surface, the greater the difference in grayscale value between the crack area and the normal wall surface.

[0036] Therefore, the collected grayscale image of the basement exterior wall surface is binarized to obtain a binary image.

[0037] It should be noted that there are many commonly used binarization processing methods. In this embodiment, the grayscale image of the basement exterior wall surface is used as the input of the Otsu threshold segmentation algorithm, and the segmentation threshold is output. The grayscale values ​​corresponding to the pixels whose grayscale values ​​are greater than or equal to the segmentation threshold are set to 0, and the grayscale values ​​corresponding to the pixels whose grayscale values ​​are less than the segmentation threshold are set to 1. The implementer may also adopt other binarization processing methods. This embodiment does not impose any special restrictions on the selection of the binarization processing method.

[0038] (4) Further, the result of bitwise multiplication of the grayscale image of each basement exterior wall surface and the binary image is used as the target image, wherein the bitwise multiplication refers to multiplying all elements in the grayscale image of the basement exterior wall surface with the pixel values ​​at the corresponding same position in the binary image, and the specific calculation steps are not repeated here.

[0039] (5) Furthermore, a connected domain extraction algorithm is used to obtain all connected domains in the target image, and the OpenCV library is used to obtain the minimum bounding rectangle of each connected domain.

[0040] In addition, it should be understood that there are many commonly used connected domain extraction algorithms. In this embodiment, a union-find algorithm is used to extract connected domains in the target image. The implementer may also use other connected domain extraction algorithms such as morphology-based algorithms. This embodiment does not impose any special restrictions on the selection of connected domain extraction algorithms.

[0041] Among them, Otsu threshold segmentation algorithm, union-find algorithm and OpenCV library are all well-known technologies. The specific process of binarizing the image using Otsu threshold segmentation algorithm, the specific process of extracting the connected domain in the image using the union-find algorithm, and the specific process of obtaining the minimum circumscribed rectangle of the connected domain using the OpenCV library will not be described in detail.

[0042] Preferably, the schematic diagram of underground structure exterior wall image acquisition and processing process provided in this embodiment is as follows: Figure 2 shown.

[0043] Step S2: Determine the morphological anomaly index of each connected domain according to the aspect ratio of the minimum circumscribed rectangle of each connected domain and the angle between the left edge of the minimum circumscribed rectangle and the horizontal line where the bottom edge of the basement outer wall is located.

[0044] When a large-scale basement is constructed across seasons, the basement exterior wall is exposed for a long time, and the temperature difference between the inside and outside of the basement exterior wall is large. The strength grade of the basement exterior wall concrete is high, the amount of concrete cement is high, the heat generation is large, and the self-shrinkage is large. Therefore, when the wall shrinkage is constrained by the floor, the shrinkage deformation is perpendicular to the constraint direction, and the cracks on the basement exterior wall appear in the form of a vertical structure, and the length of most cracks is close to the wall height. However, although there are many cracks on the surface of the basement exterior wall, the crack width is usually narrow, and the crack width does not change significantly over time. In addition, water stains often appear around the cracks due to structural stress, material drying shrinkage and other factors, but due to the lack of a continuous water supply, it usually does not lead to the formation of a large number of water stains, so the cracks and the surrounding water stains are usually narrow.

[0045] When the basement exterior wall shrinks and deforms, it cannot deform freely, which causes shrinkage cracks to form on the basement exterior wall. At the same time, the wall shrinkage will be constrained by the basement floor. Under this effect, the cracks on the basement exterior wall present a vertical structure.

[0046] Therefore, by analyzing the difference between the length and width of the minimum circumscribed rectangle of each connected domain, and the angle between the left edge of the minimum circumscribed rectangle and the horizontal line where the bottom edge of the basement outer wall is located, the morphological anomaly index of each connected domain is determined, specifically:

[0047] The morphological abnormality index CMA of the connected domain i i The expression is: In the formula, h i ,d i Respectively represent the length and width of the smallest external rectangle of the connected domain i; θ i Represents the angle between the left edge of the minimum circumscribed rectangle of the connected domain i and the horizontal line where the bottom edge of the basement outer wall is located; tan() represents the tangent function.

[0048] According to the morphological anomaly index of each connected domain, it can be understood that when the ratio of the length to the width of the minimum circumscribed rectangle of the connected domain is larger, the connected domain is more likely to be the area where the cracks on the surface of the basement outer wall are located, because the crack area presents a narrow and long feature, and because the cracks are usually distributed vertically, the closer the angle between the left edge of the minimum circumscribed rectangle of the connected domain and the horizontal line is to 90 degrees, the more likely the area where the connected domain is located is the crack on the surface of the basement outer wall, and the larger the final morphological anomaly index is; conversely, when the ratio of the length to the width of the minimum circumscribed rectangle of the connected domain is smaller, and the angle between the left edge of the minimum circumscribed rectangle of the connected domain and the horizontal line is closer to 90 degrees, it means that the area where the connected domain is located is less likely to be the crack on the surface of the basement outer wall, and the smaller the final morphological anomaly index is.

[0049] Step S3: Use a skeleton extraction algorithm to obtain the skeleton of each connected domain, determine the symmetry difference of each pixel point on the skeleton of each connected domain based on the distance from each pixel point on the skeleton of each connected domain to the left and right edges of the minimum circumscribed rectangle of the corresponding connected domain, and determine the irregularity index of each connected domain in combination with the degree of curvature at each pixel point on the skeleton of each connected domain; determine the crack confidence of each connected domain based on the irregularity index and the morphological anomaly index.

[0050] When cracks appear on the surface of the basement's outer wall, the cracks usually exist in an isolated state, and the crack area is usually accompanied by water seepage. The water stains will extend along the crack direction and penetrate around the cracks, resulting in the formation of a larger water stain connected domain outside the connected domain where the cracks are located on the basement's outer wall surface. And because the effect of gravity on water stain penetration is the same when water stains penetrate to the left and right sides of the cracks, and because the cracks on the basement's outer wall present a vertical structure, the connected domain formed by the water stains presents a left-right symmetrical structure with the cracks as the center.

[0051] Therefore, considering the symmetry of the connected domain where the crack is located due to the influence of water stains, the distance between the pixel points in the crack area in the connected domain and the left and right edges of the minimum circumscribed rectangle of the connected domain is analyzed to determine the symmetry difference of each pixel point on the skeleton in each connected domain, specifically:

[0052] In the crack area on the surface of the basement outer wall, since the crack itself is a concave structure, light entering it will be partially absorbed, resulting in the grayscale of the crack being lower than that of the water stain area. Therefore, this embodiment uses a skeleton extraction algorithm to extract the skeleton of each connected domain in the target image.

[0053] It should be noted that there are many commonly used contour extraction algorithms. In this embodiment, the Zhang-Suen algorithm is used to extract the skeleton of each connected domain in the binary image. The implementer may also use other skeleton extraction algorithms. Regarding the selection of the skeleton extraction algorithm, this embodiment does not impose any special restrictions.

[0054] Among them, the Zhang-Suen algorithm is a well-known technology in the field of skeleton extraction, and its specific principle is not repeated here.

[0055] Furthermore, the symmetric difference D of the pixel m on the skeleton in the connected domain i is i,m The expression is: D i,m =|l i,m -r i,m |; In the formula, l i,m 、r i,m They respectively represent the distances from the pixel point m on the skeleton within the connected domain i to the left and right edges of the minimum outer rectangle of the connected domain i.

[0056] It should be noted that, since cracks usually exist in isolated forms, the water stain area usually contains only one crack. Therefore, when the connected domain may be the connected domain where the water stain and the crack are located, the connected domain only corresponds to the skeleton of the crack. When the connected domain is not the connected domain where the water stain and the crack are located, there is generally no skeleton structure in the corresponding connected domain. In particular, if there is no skeleton in the connected domain, the connected domain will not be analyzed.

[0057] Furthermore, according to the symmetric difference of each pixel point on the skeleton in each connected domain, it can be understood that the smaller the difference in distance from the pixel point on the skeleton in the connected domain to the left and right edges of the minimum circumscribed rectangle of the connected domain, the smaller the symmetric difference, which means it is more likely to be the area where the crack is located; conversely, the greater the difference in distance from the pixel point on the skeleton in the connected domain to the left and right edges of the minimum circumscribed rectangle of the connected domain, the greater the symmetric difference, which means it is less likely to be the area where the crack is located.

[0058] In addition, there are regular horizontal and vertical grooves on the surface of the basement's outer wall, and the grooves serve as drainage channels on the wall, guiding the water on the wall to the drainage outlet, making the longitudinal grooves very similar to wall cracks in structure. However, the edges of the grooves are flat and smooth, while the cracks on the basement's outer wall are caused by the different stresses on the various parts of the wall surface, which leads to the destruction of the wall structure. Therefore, the grooves on the basement's outer wall have regular shapes and sizes in structure, and the edges of the grooves are flat and smooth, while the cracks on the basement's outer wall have irregular shapes and sizes, and the edges of the cracks are relatively rough.

[0059] Therefore, by analyzing the degree of curvature at each pixel point on the skeleton of each connected domain, the irregularity index of each connected domain is determined; based on the irregularity index and the morphological anomaly index, the crack confidence of each connected domain is determined, specifically:

[0060] Irregularity index CU of connected domain i i The expression is: In the formula, q i,m represents the curvature of pixel m on the skeleton within the connected domain i; M i Represents the number of all pixel points on the skeleton within the connected domain i; β represents a preset constant greater than 0, which is used to prevent the denominator from being 0, wherein the value of β is artificially set. In this embodiment, the value of β is 0.01, and the implementer can also set it by himself according to the specific situation. This embodiment does not impose any special restrictions.

[0061] Further, according to the irregularity index of each connected domain, it can be understood that when the connected domain in the grayscale image of the basement exterior wall surface is a crack area, since the cracks on the basement exterior wall surface present an irregular shape, the larger the curvature at the pixel points on the skeleton of the connected domain, and due to the water seepage phenomenon of the cracks on the basement exterior wall surface, the connected domain formed by the water stains presents a left-right symmetrical structure with the cracks on the exterior wall surface as the center, so the smaller the distance difference from the pixel points on the skeleton of the connected domain to the left and right edges of the minimum circumscribed rectangle of the connected domain, that is, the smaller the symmetry difference, the smaller the curvature at the pixel points on the skeleton of the connected domain, and thus the larger the irregularity index of the obtained connected domain; conversely, if the curvature at the pixel points on the skeleton of the connected domain is smaller, the larger the distance difference from the pixel points on the skeleton of the connected domain to the left and right edges of the minimum circumscribed rectangle of the connected domain, that is, the larger the symmetry difference, the smaller the irregularity index of the obtained connected domain, which means that it is less likely that the current connected domain is the crack area.

[0062] Further, based on the irregularity index and the morphological anomaly index, the crack confidence of each connected domain is determined, specifically:

[0063] The fracture confidence of each connected domain is the product of the irregularity index and the morphological anomaly index of each connected domain.

[0064] According to the fracture confidence of each connected domain, it can be understood that when the morphological anomaly index of the connected domain is larger and the irregularity index is larger, the fracture confidence of the connected domain is larger, indicating that the current connected domain is more likely to be the fracture area; conversely, when the morphological anomaly index of the connected domain is smaller and the irregularity index is smaller, the fracture confidence of the connected domain is smaller, indicating that the current connected domain is less likely to be the fracture area.

[0065] Preferably, the schematic diagram of the crack confidence extraction process provided in this embodiment is as follows: Figure 3 shown.

[0066] Step S4: Based on the average distribution of crack confidences of all connected domains in the target image, determine the crack defect degree of each basement exterior wall surface; based on the distribution of crack defect degrees of all basement exterior wall surfaces, determine the crack assessment threshold and detect the basement exterior wall surface to be detected.

[0067] According to the analysis of steps S2-S3, the crack confidence of each connected domain in the grayscale image of the basement exterior wall surface can be obtained. The crack confidence comprehensively considers the growth characteristics of the crack area and the water seepage phenomenon of the cracks on the basement exterior wall surface. Therefore, based on the crack confidence, it is possible to accurately detect whether there are crack defects on the basement exterior wall surface.

[0068] The average of the crack confidences of all connected domains in the target image of each basement outer wall surface is taken as the crack defect degree of each basement outer wall surface;

[0069] Furthermore, the crack assessment threshold is the average of the crack defect degrees of a preset number of basement exterior wall surfaces;

[0070] A grayscale image and a binary image of the exterior wall surface of the basement to be inspected are obtained, and the crack defect degree of the exterior wall surface of the basement to be inspected is obtained according to the calculation method of the crack defect degree of each outdoor wall surface in steps S1-S3. If the crack defect degree of the exterior wall surface of the basement to be inspected is greater than the crack assessment threshold, then there is a crack defect on the exterior wall surface of the basement to be inspected; otherwise, there is no crack defect on the exterior wall surface of the basement to be inspected.

[0071] Based on the same inventive concept as the above-mentioned method, an embodiment of the present application also provides a visual detection system for underground structure exterior wall defects, including a memory, a processor, and a computer program stored in the memory and running on the processor, and when the processor executes the computer program, the steps of any one of the above-mentioned methods for visual detection of underground structure exterior wall defects are implemented.

[0072] It should be noted that the above sequence of the embodiments of the present application is for description only and does not represent the advantages and disadvantages of the embodiments. The above is a description of a specific embodiment of this specification. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0073] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

[0074] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present application should be included in the protection scope of the present application.

Claims

1. A method for visually detecting defects in the outer wall of an underground structure, characterized in that: The method comprises the following steps: Obtain the grayscale image and binary image of the outer wall surface of each basement, use the result of bitwise multiplication of the grayscale image and the binary image as the target image, extract all connected domains in the target image, and obtain the minimum circumscribed rectangle of each connected domain; According to the aspect ratio of the minimum circumscribed rectangle of each connected domain and the angle between the left edge of the minimum circumscribed rectangle and the horizontal line where the bottom edge of the basement outer wall is located, the morphological anomaly index of each connected domain is determined; A skeleton extraction algorithm is used to obtain the skeleton in each connected domain. Based on the distance from each pixel point on the skeleton in each connected domain to the left and right edges of the minimum circumscribed rectangle of the corresponding connected domain, the symmetry difference of each pixel point on the skeleton in each connected domain is determined. Combined with the degree of curvature at each pixel point on the skeleton of each connected domain, the irregularity index of each connected domain is determined. Based on the irregularity index and the morphological anomaly index, the crack confidence of each connected domain is determined. Based on the average distribution of crack confidences of all connected domains in the target image, the crack defect degree of each basement exterior wall surface is determined; based on the distribution of crack defect degrees of all basement exterior wall surfaces, the crack assessment threshold is determined, and the basement exterior wall surface to be detected is detected.

2. A method for visually detecting defects in an underground structure exterior wall as claimed in claim 1, characterized in that: The expression of the morphological abnormality index of each connected domain is: In the formula, CMA i represents the morphological abnormality index of the connected domain i; h i ,d i Respectively represent the length and width of the smallest external rectangle of the connected domain i; θ i Represents the angle between the left edge of the minimum circumscribed rectangle of the connected domain i and the horizontal line where the bottom edge of the basement outer wall is located; tan() represents the tangent function.

3. A method for visually detecting defects in an underground structure exterior wall as claimed in claim 1, characterized in that: The skeleton extraction algorithm is the Zhang Suen thinning algorithm.

4. A method for visually detecting defects in an underground structure exterior wall as claimed in claim 1, characterized in that: The expression of the symmetric difference degree of each pixel point on the skeleton in each connected domain is: i,m =|l i,m -r i,m |; In the formula, D i,m represents the symmetric difference of pixel point m on the skeleton within the connected domain i; l i,m 、r i,m They respectively represent the distances from the pixel point m on the skeleton within the connected domain i to the left and right edges of the minimum outer rectangle of the connected domain i.

5. A method for visually detecting defects in an underground structure exterior wall as claimed in claim 4, characterized in that: The expression of the irregular index of each connected domain is: Where, CU i represents the irregular index of the connected domain i; q i,m represents the curvature of pixel m on the skeleton within the connected domain i; M i Represents the number of all pixels on the skeleton within the connected domain i; β represents a constant that is preset to be greater than 0.

6. A method for visually detecting defects in an underground structure exterior wall as claimed in claim 1, characterized in that: The crack confidence of each connected domain is the product of the irregularity index and the morphological anomaly index of each connected domain.

7. A method for visually detecting defects in an underground structure exterior wall as claimed in claim 1, characterized in that: The crack defect degree of each basement outer wall surface is the average of the crack confidences of all connected domains in the target image of each basement outer wall surface.

8. A method for visually detecting defects in an underground structure exterior wall as claimed in claim 1, characterized in that: The crack assessment threshold is the average value of the crack defect degrees of a preset number of basement exterior wall surfaces.

9. A method for visually detecting defects in an underground structure exterior wall as claimed in claim 1, characterized in that: The detection of the surface of the outer wall of the basement to be detected includes: A grayscale image and a binary image of the exterior wall surface of the basement to be inspected are obtained, and the crack defect degree of the exterior wall surface of the basement to be inspected is obtained according to the calculation method of the crack defect degree of each outdoor wall surface. If the crack defect degree of the exterior wall surface of the basement to be inspected is greater than the crack assessment threshold, then there is a crack defect on the exterior wall surface of the basement to be inspected; otherwise, there is no crack defect on the exterior wall surface of the basement to be inspected.

10. A visual inspection system for defects in an underground structure exterior wall, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the method for visually detecting defects in the exterior wall of an underground structure as described in any one of claims 1-9 are implemented.

Images