A municipal bridge damage defect detection method and system

Abstract

The present invention relates to the field of bridge damage detection, and more specifically, the present invention relates to a method and system for detecting damage defects in municipal bridges. The method comprises: performing edge detection and connected domain extraction on a collected bridge surface grayscale image to obtain a number of edge pixel points and a number of connected domains, and calculating the regularity of a target area; constructing a grayscale run matrix of a target rectangle, and calculating the texture disorder degree of the target area according to the grayscale run matrix; weighted summing and normalizing the regularity and texture disorder degree, and using the normalized result as the initial abnormality of the bridge; emitting an ultrasonic pulse to the target area to obtain the echo time and echo intensity, and using the echo time and echo intensity as a correction factor to correct the initial abnormality of the bridge to obtain a bridge damage index, and completing defect detection. Through the technical solution of the present invention, the efficiency and accuracy of bridge defect detection can be improved.

Description

Technical Field

[0001] The present invention relates to the field of bridge damage detection, and more specifically, to a municipal bridge damage defect detection method and system. Background Art

[0002] Municipal bridges, as a key component of urban infrastructure, undertake the important task of connecting various areas of the city and facilitating traffic flow. Bridge cracks are one of the most common bridge damages. They may be caused by a variety of factors, such as vehicle overloading causing structural stress to exceed the design limit, material aging causing concrete and steel to lose their original strength and toughness, and environmental erosion such as acid rain and chloride ion intrusion causing steel corrosion. The appearance of these cracks will not only reduce the bearing capacity of the bridge and shorten its service life, but may also cause more serious structural damage due to the expansion and deepening of the cracks, such as partial or overall collapse, which will not only cause significant casualties, but may also lead to traffic interruptions, economic losses and social disorder. Therefore, monitoring, evaluating and timely repairing bridge cracks are important measures to ensure the safe operation of bridges and extend their service life.

[0003] The existing Chinese patent application document with publication number CN114839192A discloses a concrete bridge crack detection system, which includes collecting images through an acquisition device; image preprocessing; characterizing the cracks according to the minimum grayscale value existing in the crack arrangement area to realize CCPM method crack edge detection; using the projection method to make an initial judgment on the cracks; and judging the crack image type.

[0004] However, since there may be some textures or stains on the bridge surface, these will affect the accuracy of the edge detection results, causing many interference edges to appear in the edge detection results, affecting the subsequent analysis of the crack edges, resulting in a reduction in the accuracy of the bridge crack detection results. Summary of the invention

[0005] In order to solve the problem of reduced accuracy of bridge crack detection results, the present invention proposes a municipal bridge damage defect detection method and system.

[0006] In a first aspect, the present invention discloses a method for detecting damage defects of a municipal bridge, comprising: performing edge detection and connected domain extraction on a collected bridge surface grayscale image to obtain a number of edge pixel points and a number of connected domains, taking any connected domain as a target area, and calculating the regularity of the target area; taking the minimum circumscribed rectangle of the target area as a target rectangle, constructing a grayscale run-length matrix of the target rectangle, and calculating the texture disorder degree of the target area according to the grayscale run-length matrix; weighted summing and normalizing the regularity and texture disorder degree, and taking the normalized result as the initial abnormality of the bridge; emitting an ultrasonic pulse to the target area to obtain an echo time and an echo intensity, and taking the echo time and the echo intensity as correction factors to correct the initial abnormality of the bridge to obtain a bridge damage index, and completing defect detection; wherein the bridge damage index satisfies the relationship: , represents the bridge damage index, represents the initial abnormality of the bridge, Indicates the maximum echo time, represents the ideal echo time, Indicates the actual echo time, Indicates the echo strength, Represents the normalization function.

[0007] By combining bridge surface image analysis with ultrasonic detection technology, a comprehensive assessment of bridge damage can be conducted from two dimensions: morphological characteristics and texture characteristics, which can effectively identify and quantify the extent of surface cracks and weathering erosion.

[0008] Preferably, the regularity includes: performing skeleton extraction on the target area to obtain a center line, and taking the pixel points on the center line as the center line points; the regularity satisfies the relationship: , Indicates regularity, Indicates the centerline point The curvature of Indicates the number of centerline points of the target area, represents the area of ​​the target region, Represents an exponential function.

[0009] The cumulative and average values ​​of the curvature are adjusted in the form of an exponential function, which can effectively reveal the degree of shape change of the region, especially the morphological anomalies in cracks or weathering and erosion areas.

[0010] Preferably, the regularity further comprises: obtaining corner points of the target area by corner point detection; the regularity satisfies the relationship: , Indicates regularity, Indicates the number of corner points whose internal angle is greater than the straight angle. Represents the total number of corner points, Indicates The internal angle of the corner point, Represents the normalization function.

[0011] Paying attention to the number of corner points whose internal angles are greater than the straight angle can reflect the complexity and irregularity of the shape of the target area. When the corner points of the target area show higher internal angle values ​​and more irregular corner points, it indicates that the shape of the area is greatly deformed or damaged. The regularity index can quantify the regularity of the area by weighted summing and normalizing the internal angles of the corner points.

[0012] Preferably, the texture disorder degree satisfies the relationship:

[0013] , Indicates the degree of texture disorder. Represents the grayscale run matrix Line The element value of the column, represents the probability, Represents a logarithmic function.

[0014] By calculating the grayscale run-length matrix to evaluate the texture disorder of the target area, the details and complexity of the crack area can be effectively quantified, thereby providing texture support for bridge damage detection. The grayscale run-length matrix reflects the distribution of grayscale values ​​and run-length distances in the image, thereby revealing the texture characteristics and structural complexity of the crack area.

[0015] Preferably, the texture disorder degree also satisfies the relationship:

[0016] , Indicates the degree of texture disorder. Represents the grayscale run matrix Line The element value of the column.

[0017] Preferably, the weighting uses an optimal weight.

[0018] Preferably, obtaining the optimal weight includes: presetting weights, calculating a historical bridge damage index based on historical bridge data, and using the calculated historical bridge damage index as a predicted value; judging an actual damage index of historical bridge data based on a bridge damage grade, and using the actual damage index as an actual value; calculating a loss value between a predicted value and an actual value using a cross entropy loss function, traversing the weight value, and obtaining the optimal weight when the loss value is minimized.

[0019] By comparing the difference between the predicted and actual values ​​and minimizing the loss value, the weights can be gradually adjusted to ensure that the model can accurately reflect the damage condition of the actual bridge. This process not only enhances the reliability and accuracy of the prediction model, but also improves the accuracy of the damage index prediction by optimizing the weights.

[0020] In a second aspect, the present invention discloses a municipal bridge damage defect detection system, comprising: a processor; and a memory, wherein the memory stores computer instructions, and when the computer instructions are executed by the processor, the system executes the above-mentioned municipal bridge damage defect detection method.

[0021] Beneficial effects of the present invention:

[0022] The present invention performs edge detection and connected domain extraction on the grayscale image of the bridge surface, calculates the regularity and texture disorder of the target area, and thus evaluates the damage of the bridge surface from the two dimensions of shape and texture. The calculation of regularity involves skeleton extraction and curvature analysis, which can reveal the geometric shape anomalies of the crack area and the weathering erosion area, especially the changes in the width, depth and morphological characteristics of the cracks, and provides a detailed identification basis; while the calculation of the texture disorder degree is based on the grayscale run matrix, which reflects the texture changes of the crack area and the weathering erosion area, and reflects the degree of local structural disorder in the damaged area. After weighted summation and normalization, these features can more scientifically integrate the influence of different damage types, and the optimal selection of weights further improves the accuracy and reliability of the evaluation. Combined with the determination of the optimal weights, the calculation of the initial abnormality of the bridge is more scientific and accurate. The introduction of ultrasonic detection further distinguishes surface cracks and structural cracks from a three-dimensional direction, and improves the depth and accuracy of detection. Finally, through the calculation of the bridge damage index, the image and ultrasonic data can be integrated to provide a quantitative evaluation of bridge damage, thereby realizing efficient and accurate detection of bridge damage and ensuring the safety and reliability of the bridge structure. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] By reading the following detailed description with reference to the accompanying drawings, the above and other objects, features and advantages of the exemplary embodiments of the present invention will become readily understood. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein:

[0024] Figure 1 It is a flow chart of a municipal bridge damage defect detection method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0025] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0026] It should be understood that when the terms "first", "second", etc. are used in the claims, descriptions, and drawings of the present invention, they are only used to distinguish different objects, rather than to describe a specific order. The terms "include" and "comprise" used in the description and claims of the present invention indicate the presence of the described features, wholes, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their collections.

[0027] The present invention provides a method for detecting damage and defects of municipal bridges. Figure 1 As shown, a municipal bridge damage defect detection method includes steps S1 to S4, which are described in detail below.

[0028] S1, edge detection and connected domain extraction are performed on the collected bridge surface grayscale image to obtain a number of edge pixels and a number of connected domains, and the regularity of the target area is calculated.

[0029] In one embodiment, a high-definition camera mounted on a drone is used to photograph a municipal bridge to obtain a surface RGB (Red Green Blue) image, and the surface RGB image is grayed to obtain an initial bridge surface grayscale image, and the bridge surface grayscale image is Gaussian filtered and histogram equalized to obtain a bridge surface grayscale image. Gaussian filtering and histogram equalization are used to remove noise from the initial bridge surface grayscale image and enhance the contrast of the initial bridge surface grayscale image to make it clearer.

[0030] It should be noted that since bridges are exposed to various external environmental conditions for a long time (such as climate change, traffic loads, etc.), they are prone to various forms of damage. After edge detection of the bridge grayscale image, in addition to possible crack contours, there are also edges formed by textures or stains on the bridge surface itself.

[0031] Edge detection and connected domain extraction of the grayscale image of the bridge surface can obtain several connected domains. The connected domains are crack areas or weathering and erosion areas. Any connected domain is taken as the target area, the regularity of the target area is calculated, the skeleton of the target area is extracted to obtain the center line, and the pixel points on the center line are taken as the center line points.

[0032] The regularity satisfies the relationship:

[0033] , Indicates regularity, Indicates the centerline point The curvature of Indicates the number of centerline points of the target area, represents the area of ​​the target region, Represents an exponential function.

[0034] The area of ​​the target area is usually expressed by the number of pixels. The larger the ratio of the area to the number of centerline points, the fewer the number of centerline points and the shorter the centerline. When the area of ​​the target area remains unchanged, the overall shape of the target area tends to be elliptical or circular. At this time, the target area is more likely to be a weathering and erosion area, so the regularity is smaller. Smaller curvature values ​​usually indicate that the centerline is smoother and has higher regularity.

[0035] It should be noted that the crack area is usually a structural fracture caused by crustal stress. These cracks may appear as straight lines or regular shapes, forming relatively neat geometric features. Cracks are often distributed along specific directions and patterns, showing a high degree of regularity. For example, geological structures such as faults and folds often have relatively regular shapes, and their cracks show the characteristics of consistent directions and uniform spacing.

[0036] Weathering erosion is a surface geological process, which is mainly caused by the gradual breaking, erosion and transportation of the surface of rocks or soil under the action of physical, chemical, wind, water and other natural factors. The weathering erosion process is usually complicated and affected by many factors, including climate, lithology, topography, time, etc. Due to the randomness and complexity of these factors, the landforms formed by weathering erosion are generally irregular. On bridges, the areas produced by weathering erosion are irregular, and the surface shows different degrees of roughness, erosion and unevenness, so the regularity is low.

[0037] In one embodiment, the regularity further includes: obtaining corner points of the target area by using corner point detection.

[0038] The regularity satisfies the relationship: , Indicates regularity, Indicates the number of corner points whose internal angle is greater than the straight angle. Represents the total number of corner points, Indicates The internal angle of the corner points, Represents the normalization function.

[0039] It should be noted that if the internal angle is greater than a straight angle (i.e. 180 degrees), it means that the direction of the edge at the corner point changes dramatically, forming an inward bend.

[0040] S2, construct the grayscale run-length matrix of the target rectangle, and calculate the texture disorder degree of the target area according to the grayscale run-length matrix.

[0041] Calculating the texture disorder of the target area based on the grayscale run-length matrix can effectively quantify the details and complexity of the crack area, thereby identifying the damage of the bridge. In crack detection, the boundary of the crack is usually located by the minimum enclosing rectangle, and then the run-length matrix is ​​calculated based on the grayscale value of the area to analyze the texture features.

[0042] In one embodiment, the texture disorder degree satisfies the relationship:

[0043] , Indicates the degree of texture disorder. Represents the grayscale run matrix Line The element value of the column, represents the probability, Represents a logarithmic function.

[0044] The meaning of the element value is the gray value , the travel distance is Number of occurrences.

[0045] By calculating the grayscale run-length matrix to evaluate the texture disorder of the target area, the details and complexity of the crack area can be effectively quantified, thereby providing texture support for bridge damage detection. The grayscale run-length matrix reflects the distribution of grayscale values ​​and run-length distances in the image, thereby revealing the texture characteristics and structural complexity of the crack area.

[0046] In one embodiment, the texture disorder degree also satisfies the relationship:

[0047] , Indicates the degree of texture disorder. Represents the grayscale run matrix Line The element value of the column.

[0048] S3, weighted sum of regularity and texture disorder degree and normalize them, and use the normalized result as the initial abnormality of the bridge.

[0049] In one embodiment, the optimal weight is used for the weighting of regularity and texture disorder, and obtaining the optimal weight includes: presetting weights, calculating the historical bridge damage index based on the historical bridge data, and using the calculated historical bridge damage index as the predicted value; judging the actual damage index of the historical bridge data based on the bridge damage level, and using the actual damage index as the actual value; using the cross entropy loss function to calculate the loss value of the predicted value and the actual value, traversing the value of the weight, and obtaining the optimal weight when the loss value is the smallest.

[0050] It should be noted that according to the article "Research on Fuzzy Assessment of Overall Damage Degree of Bridges Based on Stiffness Damage Index" in the Journal Network of the Chinese Society of Mechanics, the article number is 1000-4750 (2011) 12-0092-06. The article gives a description of the damage level of bridges and the range of damage index. The damage index between 0 and 0.2 is basically intact and does not need to be repaired; the damage index between 0.2 and 0.4 is slightly damaged and requires minor repairs; the damage index between 0.4 and 0.6 is moderately damaged and requires medium repairs; the damage index between 0.6 and 0.9 is severely damaged and requires major repairs; the damage index between 0.9 and 1 is collapsed and cannot be repaired. The damage index range given in the article is the actual damage index.

[0051] In one embodiment, weights are preset for the regularity and the texture disorder degree respectively, and the result of weighted sum normalization is used as the initial abnormality of the bridge.

[0052] S4, transmitting ultrasonic pulses to the target area to obtain the echo time and echo intensity, using the echo time and echo intensity as correction factors to correct the initial abnormality of the bridge to obtain the bridge damage index, and completing the defect detection.

[0053] Ultrasonic testing technology is a non-destructive testing method commonly used for bridge structural health monitoring. It can effectively detect cracks, cavities, corrosion, fatigue and other damage defects in bridges. This method uses the propagation characteristics of sound waves in bridge materials to detect abnormal changes in the structure.

[0054] It should be noted that ultrasonic testing technology can be used to determine surface cracks and structural cracks. Surface cracks are generally shallow, and the wave reflection time is short. Since cracks usually only affect the local wave speed, the echo time is relatively fast.

[0055] For surface cracks, the echo intensity is usually weak. This is because surface cracks usually cause part of the wave energy to be reflected or scattered. The crack surface is relatively smooth and may not reflect the echo strongly, so the echo intensity is low. Structural cracks are often deeper than surface cracks and may extend to the inside of the bridge structure. Deeper cracks can more significantly reflect or scatter ultrasonic waves, so the echo intensity is relatively high. Especially when cracks are formed in deeper structures, ultrasonic waves may be reflected multiple times, resulting in stronger echo signals.

[0056] In one embodiment, the bridge damage index satisfies the relationship:

[0057] , represents the bridge damage index, represents the initial abnormality of the bridge, Indicates the maximum echo time, represents the ideal echo time, Indicates the actual echo time, Indicates the echo strength, Represents the normalization function.

[0058] The maximum echo time represents the most serious defect situation, and the ideal echo time represents the echo time when there is no defect.

[0059] The bridge damage index of each connected domain is traversed, the mean of all bridge damage indices is taken as the final damage result, the final damage result and the damage index range are compared to complete the municipal bridge damage defect detection.

[0060] An embodiment of the present invention further discloses a municipal bridge damage defect detection system, including a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a municipal bridge damage defect detection method according to the present invention is implemented.

[0061] The above system also includes other components well known to those skilled in the art, such as a communication bus and a communication interface, and their configuration and functions are known in the art, so they will not be described in detail here.

[0062] In the present invention, the aforementioned memory may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system, apparatus or device. For example, a computer-readable storage medium may be any appropriate magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory RRAM (Resistive Random Access Memory), a dynamic random access memory DRAM (Dynamic Random Access Memory), a static random access memory SRAM (Static Random-Access Memory), an enhanced dynamic random access memory EDRAM (Enhanced Dynamic Random Access Memory), a high-bandwidth memory HBM (High-Bandwidth Memory), a hybrid memory cube HMC (Hybrid Memory Cube), etc., or any other medium that can be used to store the required information and can be accessed by an application, a module or both. Any such computer storage medium may be part of a device or accessible or connectable to a device.

[0063] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, various alternatives to the embodiments of the present invention described herein may be employed.

[0064] The above are all preferred embodiments of the present invention, and are not intended to limit the protection scope of the present invention. Therefore, any equivalent changes made based on the structure, shape, and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A municipal bridge damage defect detection method, characterized in that: include: Perform edge detection and connected domain extraction on the collected bridge surface grayscale image to obtain several edge pixel points and several connected domains. Take any connected domain as the target area and calculate the regularity of the target area. The minimum circumscribed rectangle of the target area is taken as the target rectangle, the grayscale run-length matrix of the target rectangle is constructed, and the texture disorder degree of the target area is calculated according to the grayscale run-length matrix; The regularity and texture disorder are weighted and summed and normalized, and the normalized result is used as the initial abnormality of the bridge; Ultrasonic pulses are emitted to the target area to obtain the echo time and echo intensity. The echo time and echo intensity are used as correction factors to correct the initial abnormality of the bridge to obtain the bridge damage index and complete the defect detection. Among them, the bridge damage index satisfies the relationship: , represents the bridge damage index, represents the initial abnormality of the bridge, Indicates the maximum echo time, represents the ideal echo time, Indicates the actual echo time, Indicates the echo strength, Represents the normalization function.

2. A municipal bridge damage defect detection method according to claim 1, characterized in that: The regularity includes: Perform skeleton extraction on the target area to obtain the center line, and use the pixel points on the center line as the center line points; The regularity satisfies the relationship: , Indicates regularity, Indicates the centerline point The curvature of Indicates the number of centerline points of the target area, represents the area of ​​the target region, Represents an exponential function.

3. A municipal bridge damage defect detection method according to claim 1, characterized in that: The regularity also includes: Use corner point detection to obtain the corner points of the target area; The regularity satisfies the relationship: , Indicates regularity, Indicates the number of corner points whose internal angle is greater than the straight angle. Represents the total number of corner points, Indicates The internal angle of the corner points, Represents the normalization function.

4. A municipal bridge damage defect detection method according to claim 1, characterized in that: The texture disorder degree satisfies the relationship: , Indicates the degree of texture disorder. Represents the grayscale run matrix Line The element value of the column, represents the probability, Represents a logarithmic function.

5. A municipal bridge damage defect detection method according to claim 1, characterized in that: The texture disorder degree also satisfies the relationship: , Indicates the degree of texture disorder. Represents the grayscale run matrix Line The element value of the column.

6. A municipal bridge damage defect detection method according to claim 1, characterized in that: The weighting uses the optimal weight.

7. A municipal bridge damage defect detection method according to claim 6, characterized in that: Obtaining the optimal weight includes: Preset weights, calculate historical bridge damage index based on historical bridge data, and use the calculated historical bridge damage index as a predicted value; According to the bridge damage level, the actual damage index of the bridge historical data is determined, and the actual damage index is taken as the actual value; The cross entropy loss function is used to calculate the loss value of the predicted value and the actual value, and the weight value is traversed. When the loss value is minimized, the optimal weight is obtained.

8. A municipal bridge damage defect detection system, characterized in that: include: Processor; and A memory storing computer instructions, wherein when the computer instructions are executed by a processor, the system executes a municipal bridge damage defect detection method according to any one of claims 1-7.

Images