Method and system for identifying water damage of asphalt pavement

Abstract

The invention discloses an asphalt pavement water damage identification method and system. The method comprises the following steps: judging whether each correlation degree is greater than a first preset threshold value or not; if a certain association degree is not greater than a first preset threshold value, removing the first asphalt pavement image from a certain asphalt pavement image set, and updating the certain asphalt pavement image set after the first asphalt pavement image is removed according to a preset updating strategy to obtain at least one updated target asphalt pavement image set; selecting a certain target asphalt pavement image from a certain target asphalt pavement image set, acquiring a target core sample image at the position of a second target disease area in the certain target asphalt pavement image, and identifying the target core sample image, and taking the identification result as the identification result of all target asphalt pavement images in a certain target asphalt pavement image set. The water damage condition of the asphalt pavement can be accurately identified while core sample collection of the asphalt pavement is reduced.

Description

Technical Field

[0001] The present invention belongs to the technical field of asphalt pavement identification, and particularly relates to a method and system for identifying water damage of asphalt pavement. Background Art

[0002] Asphalt pavement water damage refers to the phenomenon that after water enters the asphalt pavement structure, under the action of vehicle load and temperature change, the adhesion between asphalt and aggregates decreases, and diseases such as looseness, peeling, and potholes appear on the pavement. There are differences between the repair of water damage and the repair of other non-water damage. For example, the repair of water damage includes: spraying emulsified asphalt or modified emulsified asphalt on the pavement surface to form a waterproof layer to reduce water infiltration; laying a slurry seal to repair microcracks and enhance the water tightness of the pavement. The repair of other non-water damage includes: for cracks with a width of 2 - 5 mm: filling the cracks with hot asphalt; for cracks with a width greater than 5 mm: filling the cracks with modified asphalt. Before filling the cracks, debris and fragments in the cracks need to be removed to ensure dryness. After filling the cracks, coarse sand or 3 - 5 mm stone chips are scattered on the surface to enhance the sealing effect.

[0003] In the prior art, in order to distinguish water damage well, usually core samples are extracted from the disease areas, and the captured core sample images are identified, and the water damage areas existing in the asphalt pavement can be identified more accurately. However, extracting core samples one by one for each disease area often results in a large workload, thus affecting the identification efficiency of water damage. Summary of the Invention

[0004] The present invention provides a method and system for identifying water damage of asphalt pavement to solve the technical problem that extracting core samples one by one for each disease area often results in a large workload, thus affecting the identification efficiency of water damage.

[0005] In a first aspect, the present invention provides a method for identifying water damage of asphalt pavement, including: Obtaining asphalt pavement images of a continuous section of asphalt pavement, and inputting each asphalt pavement image into a preset asphalt image recognition model, and the asphalt image recognition model outputs the disease areas in each asphalt pavement image; Dividing each asphalt pavement image according to a preset area division rule according to the positions of the disease areas in the asphalt pavement image to obtain at least one asphalt pavement image set; Selecting a first asphalt pavement image and a second asphalt pavement image in a certain asphalt pavement image set, and determining whether the first asphalt pavement image and the second asphalt pavement image are the same asphalt pavement image, where the first asphalt pavement image is the asphalt pavement image with the largest number of disease areas, and the second asphalt pavement image is the asphalt pavement image with the smallest number of disease areas; If they are not the same asphalt pavement image, determine the degree of association between the first target disease area and other disease areas in the first asphalt pavement image, and determine whether each degree of association is greater than a first preset threshold, where the first target disease area is the disease area that divides the first asphalt pavement image into the certain asphalt pavement image set, and the other disease areas are the disease areas in the first asphalt pavement image excluding the target disease area; If a certain degree of association is not greater than the first preset threshold, remove the first asphalt pavement image from the certain asphalt pavement image set, and update the certain asphalt pavement image set after removing the first asphalt pavement image according to a preset update strategy to obtain at least one updated target asphalt pavement image set; Select a certain target asphalt pavement image from a certain target asphalt pavement image set, obtain a target core sample image at the position of the second target disease area in the certain target asphalt pavement image, identify the target core sample image, and use the identification result as the identification result of all target asphalt pavement images in the certain target asphalt pavement image set.

[0006] In a second aspect, the present invention provides an identification system for water damage of asphalt pavement, characterized by comprising: An acquisition module configured to acquire asphalt pavement images of a continuous section of asphalt pavement, and input each asphalt pavement image into a preset asphalt image recognition model, and the asphalt image recognition model outputs the disease areas in each asphalt pavement image; A division module configured to divide each asphalt pavement image according to the positions of the disease areas in the asphalt pavement image by using a preset area division rule to obtain at least one asphalt pavement image set; A first judgment module configured to select a first asphalt pavement image and a second asphalt pavement image from a certain asphalt pavement image set, and judge whether the first asphalt pavement image and the second asphalt pavement image are the same asphalt pavement image, where the first asphalt pavement image is the asphalt pavement image with the largest number of disease areas, and the second asphalt pavement image is the asphalt pavement image with the smallest number of disease areas; A second judgment module configured to, if they are not the same asphalt pavement image, determine the degree of association between the first target disease area and other disease areas in the first asphalt pavement image, and determine whether each degree of association is greater than a first preset threshold, where the first target disease area is the disease area that divides the first asphalt pavement image into the certain asphalt pavement image set, and the other disease areas are the disease areas in the first asphalt pavement image excluding the target disease area; An update module, configured to remove the first asphalt pavement image from the set of asphalt pavement images of a certain one if the degree of association between the first target disease area in the first asphalt pavement image and other disease areas is not greater than a first preset threshold, and update the set of asphalt pavement images of a certain one after removing the first asphalt pavement image according to a preset update strategy to obtain at least one updated target asphalt pavement image set; An identification module, configured to select a target asphalt pavement image from a set of target asphalt pavement images of a certain one, obtain a target core sample image of the location of the second target disease area in the target asphalt pavement image of a certain one, identify the target core sample image, and use the identification result as the identification result of all target asphalt pavement images in the set of target asphalt pavement images of a certain one.

[0007] In a third aspect, an electronic device is provided, which includes: at least one processor, and a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the steps of the method for identifying water damage of an asphalt pavement according to any embodiment of the present invention.

[0008] In a fourth aspect, the present invention further provides a computer-readable storage medium, on which a computer program is stored, and when the program instructions are executed by a processor, the processor is enabled to execute the steps of the method for identifying water damage of an asphalt pavement according to any embodiment of the present invention.

[0009] The method and system for identifying water damage of an asphalt pavement of the present application determine the degree of association between the first target disease area in the first asphalt pavement image and other disease areas, and judge whether each degree of association is greater than a first preset threshold. If the degree of association of a certain one is not greater than the first preset threshold, the first asphalt pavement image is removed from the set of asphalt pavement images of a certain one, and the set of asphalt pavement images of a certain one after removing the first asphalt pavement image is updated according to a preset update strategy to obtain at least one updated target asphalt pavement image set, and a target asphalt pavement image is selected from the set of target asphalt pavement images of a certain one, a target core sample image of the location of the second target disease area in the target asphalt pavement image of a certain one is obtained, the target core sample image is identified, and the identification result is used as the identification result of all target asphalt pavement images in the set of target asphalt pavement images of a certain one. In this way, while reducing the collection of core samples of the asphalt pavement, the water damage condition of the asphalt pavement can be accurately identified. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] To more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0011] Figure 1 It is a flowchart of a method for identifying water damage to an asphalt pavement provided by an embodiment of the present invention; Figure 2 It is a structural block diagram of a system for identifying water damage to an asphalt pavement provided by an embodiment of the present invention; Figure 3 It is a schematic structural diagram of an electronic device provided by an embodiment of the present invention. Detailed implementation manners

[0012] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0013] Please refer to Figure 1 , which shows a flowchart of a method for identifying water damage to an asphalt pavement of the present application.

[0014] As Figure 1 shown, the method for identifying water damage to an asphalt pavement specifically includes the following steps: Step S101, obtain asphalt pavement images of a continuous section of asphalt pavement, and input each asphalt pavement image into a preset asphalt image recognition model, and the asphalt image recognition model outputs the disease areas in each asphalt pavement image.

[0015] In this step, the asphalt image recognition model can be trained through the YOLOv5 network or the YOLOv8 network. Specifically, collect asphalt pavement images containing cracks and ruts, ensure that the collected image data covers different lighting conditions, angles, and disease types; use annotation tools (such as LabelImg, CVAT) to annotate the cracks and ruts in the asphalt pavement images to generate bounding boxes and disease labels, and then input the annotated asphalt pavement images and disease labels into the YOLOv5 network or the YOLOv8 network for iterative training to obtain the asphalt image recognition model.

[0016] After training the asphalt image recognition model, obtain asphalt pavement images of a continuous section of asphalt pavement, and input each asphalt pavement image into the preset asphalt image recognition model. The asphalt image recognition model outputs the disease areas in each asphalt pavement image. To better identify the disease areas caused by water damage, the present invention needs to continue to execute step S102-step S106.

[0017] Step S102: According to the positions of the respective disease areas in the asphalt pavement images, divide the respective asphalt pavement images using a preset area division rule to obtain at least one asphalt pavement image set.

[0018] In this step, splice the respective asphalt pavement images in the order of acquisition time, and place the spliced asphalt pavement images in a pre-constructed two-dimensional coordinate system. Specifically, splicing the respective asphalt pavement images in the order of acquisition time means: sorting the respective asphalt pavement images in the order of acquisition time, and overlapping one edge of two adjacent sorted asphalt pavement images with the other edge, so as to splice two adjacent asphalt pavement images. For example, for asphalt pavement image A and asphalt pavement image B, overlap the right boundary line of asphalt pavement image A with the left boundary line of asphalt pavement image B to complete the splicing of asphalt pavement image A and asphalt pavement image B. Another example is to align the lower left vertex of asphalt pavement image A with the coordinate origin of the two-dimensional coordinate system, so place the spliced asphalt pavement image A and asphalt pavement image B in the pre-constructed two-dimensional coordinate system.

[0019] Obtain the coordinate information of the center points of the regions in each disease area according to the two-dimensional coordinate system. According to each coordinate information, divide the center points of the regions using the area division rule to obtain at least one set of center points of the regions. Specifically, obtaining the coordinate information of the center points of the regions in each disease area according to the two-dimensional coordinate system can be achieved by obtaining the coordinate information of each pixel point in the diseased area in the two-dimensional coordinate system and taking the average value of the coordinate information of each pixel point as the coordinate information of the center point of the region. It can also be realized by a clustering algorithm (for example, the K-Means clustering algorithm) to determine the center points of the regions in each disease area. The determination of the center points of the regions belongs to conventional technology, so it will not be elaborated here.

[0020] Divide the asphalt pavement images corresponding to the center points of the regions in a certain set of center points of the regions into the same asphalt pavement image set, that is, at least one asphalt pavement image set is obtained, where a certain set of center points of the regions is any one of the at least one set of center points of the regions.

[0021] It should be noted that the specific area division rule is as follows: different windows are set, and all area center points are divided into different windows according to the ordinate values of the center points of each area; it is judged whether the difference between the first abscissa value of the center point of the first area and the second abscissa value of the center point of the second area in the same window is greater than the second preset threshold, where the center point of the first area and the center point of the second area are two adjacent area center points; if it is greater than the second preset threshold, the center point of the first area and the center point of the second area are not divided into the same area center point set; if it is not greater than the second preset threshold, the center point of the first area and the center point of the second area are divided into the same area center point set, that is, at least one area center set is obtained.

[0022] For example, the window is a rectangular structure, and there are window A, window B and window C. Among them, the ordinate corresponding to the upper vertex of window A along the vertical axis is 4, and the ordinate corresponding to the lower vertex of window A along the vertical axis is 0. Then all area center points with ordinate values between 0 and 4 are divided into window A. Similarly, area center points that meet the requirements can be divided into window B and window C.

[0023] After the division, it is judged whether the abscissa difference between two adjacent area center points in window A is greater than the second preset threshold. If it is greater, the two adjacent area center points are not divided into the same area center point set. If it is not greater, the two adjacent area center points are divided into the same area center point set. For example, in window A, there are area center points a, area center point b, area center point c, area center point d and area center point e in sequence.

[0024] Among them, the abscissa difference between area center point a and area center point b is not greater than the second preset threshold, the abscissa difference between area center point b and area center point c is not greater than the second preset threshold, and the abscissa difference between area center point c and area center point d is not greater than the second preset threshold. Then area center point a, area center point b and area center point c are divided into the same area center point set.

[0025] Furthermore, setting different windows includes: obtaining the coordinate information of the center points of each disease area in all asphalt pavement images; clustering at least two area center points with a distance not greater than the preset distance threshold in the same asphalt pavement image according to the coordinate information to obtain at least one target area center; selecting an asphalt pavement image with the largest number of target area centers in each asphalt pavement image and defining it as the target asphalt pavement image; setting windows according to the target area centers in the target asphalt pavement image, where the length of the window along the horizontal axis of the two-dimensional coordinate system is an infinite length, and the width of the window along the vertical axis of the two-dimensional coordinate system is twice the distance from the target area center to the upper boundary or the lower boundary of the window.

[0026] In this embodiment, since water damage mostly occurs in the wheel path of the driving lane, especially in sections with frequent heavy vehicle traffic or areas with poor drainage, such as the inner side of the superelevation section and the water accumulation area of the road shoulder, water damage generally has continuity. Therefore, by dividing the center point of the area, the asphalt pavement image is synchronously divided to obtain at least one asphalt pavement image set. In this way, it is possible to divide the asphalt pavement images with the same type of water damage together as much as possible, so that after the steps S103 - S106 of the present invention are executed, the efficiency of water damage identification can be improved.

[0027] Step S103, select a first asphalt pavement image and a second asphalt pavement image from a certain asphalt pavement image set, and determine whether the first asphalt pavement image and the second asphalt pavement image are the same asphalt pavement image, where the first asphalt pavement image is the asphalt pavement image with the largest number of disease areas, and the second asphalt pavement image is the asphalt pavement image with the smallest number of disease areas.

[0028] In this step, obtain the disease areas of each asphalt pavement image in a certain asphalt pavement image set, define the asphalt pavement image with the largest number of disease areas as the first asphalt pavement image, and define the asphalt pavement image with the smallest number of disease areas as the second asphalt pavement image. Therefore, select the first asphalt pavement image and the second asphalt pavement image from a certain asphalt pavement image set, and determine whether the first asphalt pavement image and the second asphalt pavement image are the same asphalt pavement image.

[0029] In a specific embodiment, after determining whether the first asphalt pavement image and the second asphalt pavement image are the same asphalt pavement image, if they are the same asphalt pavement image, do not remove the first asphalt pavement image from a certain asphalt pavement image set, and obtain the target core sample image of the location of the first target disease area in the first asphalt pavement image; input the target core sample image into the preset asphalt pavement water damage identification model, and the asphalt pavement water damage identification model outputs the water damage identification result of the target core sample image; use the water damage identification result as the identification result of all target asphalt pavement images in a certain asphalt pavement image set.

[0030] Step S104, if they are not the same asphalt pavement image, determine the correlation degree between the first target disease area and other disease areas in the first asphalt pavement image, and determine whether each correlation degree is greater than a first preset threshold, where the first target disease area is the disease area that divides the first asphalt pavement image into the certain asphalt pavement image set, and the other disease areas are the disease areas in the first asphalt pavement image except the target disease area.

[0031] In this step, if the asphalt pavement images are not the same, it is determined whether there is partial overlap between the target disease area and a certain disease area; if there is partial overlap, the value of the correlation degree between the target disease area and a certain disease area is directly defined as the first preset threshold; if there is no partial overlap, the target center point of the target disease area and a certain center point of a certain disease area are determined, and the distance between the target center point and a certain center point is defined as the correlation degree between the target disease area and a certain disease area.

[0032] For example, the first asphalt pavement image includes disease areas A, B, and C. According to the regional center point of disease area A, the first asphalt pavement image is classified into a certain asphalt pavement image set. Therefore, disease area A has a strong correlation with the first asphalt pavement image, and disease area A is used as the target disease area.

[0033] In a specific embodiment, after determining whether each correlation degree is greater than the first preset threshold, if each correlation degree is greater than the first preset threshold, the first asphalt pavement image is not removed from a certain asphalt pavement image set, and the target core sample image at the location of the first target disease area in the first asphalt pavement image is obtained; the target core sample image is input into a preset asphalt pavement water damage identification model, and the asphalt pavement water damage identification model outputs the water damage identification result of the target core sample image; the water damage identification result is used as the identification result of all target asphalt pavement images in a certain asphalt pavement image set.

[0034] Step S105, if a certain correlation degree is not greater than the first preset threshold, the first asphalt pavement image is removed from the certain asphalt pavement image set, and the certain asphalt pavement image set after removing the first asphalt pavement image is updated according to a preset update strategy to obtain at least one updated target asphalt pavement image set.

[0035] In this step, if a certain correlation degree is not greater than a first preset threshold, the first asphalt pavement image is removed from a certain asphalt pavement image set, and it is determined whether the distance between the third target defect area in the third asphalt pavement image and the fourth target defect area in the fourth asphalt pavement image in a certain asphalt pavement image set after the first asphalt pavement image is removed is greater than a second preset threshold, wherein the third asphalt pavement image and the fourth asphalt pavement image are two adjacent asphalt pavement images, and the third target defect area is the defect area that classifies the third asphalt pavement image into a certain asphalt pavement image set, and the fourth target defect area is the defect area that classifies the third asphalt pavement image into a certain asphalt pavement image set; if the distance between the third target defect area and the fourth target defect area is greater than the second preset threshold, the third asphalt pavement image and the fourth asphalt pavement image are respectively classified into different target asphalt pavement image sets to obtain at least one target asphalt pavement image set; if the distance between the third target defect area and the fourth target defect area is not greater than the second preset threshold, the third asphalt pavement image and the fourth asphalt pavement image are respectively classified into the same target asphalt pavement image set to obtain at least one target asphalt pavement image set.

[0036] It should be noted that the third asphalt pavement image and the fourth asphalt pavement image are only used to define two adjacent asphalt pavement images. When the second asphalt pavement image is adjacent to the fourth asphalt pavement image, the second asphalt pavement image and the third asphalt pavement image are the same asphalt pavement image. The third target disease area and the fourth target disease area are the target disease areas in the third asphalt pavement image and the fourth asphalt pavement image, respectively.

[0037] In this embodiment, if a certain correlation degree is not greater than a first preset threshold, it means that the first target defect area in the first asphalt pavement image may be part of a large area caused by non-water damage. Therefore, the first asphalt pavement image is removed from a certain asphalt pavement image set, which can ensure as much as possible that the defect areas of each asphalt pavement image in the certain asphalt pavement image set after removal are all caused by water damage. And according to the preset update strategy, the certain asphalt pavement image set after the first asphalt pavement image is removed is updated to obtain at least one updated target asphalt pavement image set, so that the defect types of each defect area of ​​each asphalt pavement image in the same target asphalt pavement image set are as similar as possible, thereby effectively improving the accuracy of asphalt pavement image classification. Therefore, when executing step S106, when the recognition result can be used as the recognition result of all target asphalt pavement images in the certain target asphalt pavement image set, it is as similar as possible to the actual recognition result of traversing and identifying all target asphalt pavement images in the certain target asphalt pavement image set.

[0038] In a specific embodiment, an asphalt pavement image set sequentially includes asphalt pavement image A, asphalt pavement image B, asphalt pavement image C, asphalt pavement image D, and asphalt pavement image E. Through judgment and analysis, it can be determined that asphalt pavement image D is the first asphalt pavement image, and the correlation degree between the first target disease area and other disease areas in asphalt pavement image D is not greater than the first preset threshold. Then, asphalt pavement image D is removed from a certain asphalt pavement image set.

[0039] After that, it is judged whether the distance between the target disease areas in asphalt pavement image C and asphalt pavement image E is greater than the second preset threshold, and this distance is determined by the distance between the regional center points of the two target disease areas.

[0040] If the distance between the two target disease areas is greater than the second preset threshold, it means that the reason for this is that there happens to be a target disease area in asphalt pavement image E, which leads to the division of asphalt pavement image A, asphalt pavement image B, asphalt pavement image C, asphalt pavement image D, and asphalt pavement image E into a certain asphalt pavement image set. When asphalt pavement image D is removed, asphalt pavement image E, which was originally irrelevant to asphalt pavement image A, asphalt pavement image B, and asphalt pavement image C, is naturally separated. Therefore, asphalt pavement image C and asphalt pavement image E are respectively divided into different target asphalt pavement image sets.

[0041] Step S106, select a certain target asphalt pavement image from a certain target asphalt pavement image set, obtain the target core sample image of the position where the second target disease area is located in the certain target asphalt pavement image, identify the target core sample image, and use the identification result as the identification result of all target asphalt pavement images in the certain target asphalt pavement image set.

[0042] In this step, the second target disease area is the disease area that divides a certain target asphalt pavement image into a certain target asphalt pavement image set; Specifically, obtain the target core sample image of the position where the target disease area is located in a certain target asphalt pavement image, where a certain target asphalt pavement image is any target asphalt pavement image in a certain target asphalt pavement image set; input the target core sample image into a preset asphalt pavement water damage identification model, and the asphalt pavement water damage identification model outputs the water damage identification result of the target core sample image; use the water damage identification result as the identification result of all target asphalt pavement images in a certain target asphalt pavement image set.

[0043] It should be noted that the asphalt pavement water damage identification model can be obtained by training a convolutional neural network (CNN). For example, image processing techniques (such as gray level co-occurrence matrix, fractal dimension, etc.) are used to extract the texture features of the target core sample image, and the texture features are input into the trained asphalt pavement water damage identification model for identification to obtain the water damage identification result.

[0044] In summary, the method of the present application determines the degree of association between the first target disease area and other disease areas in the first asphalt pavement image, and judges whether each degree of association is greater than the first preset threshold. If a certain degree of association is not greater than the first preset threshold, the first asphalt pavement image is removed from a certain asphalt pavement image set, and the certain asphalt pavement image set after removing the first asphalt pavement image is updated according to a preset update strategy to obtain at least one updated target asphalt pavement image set. Then, a certain target asphalt pavement image is selected from a certain target asphalt pavement image set, the target core sample image at the position of the second target disease area in the certain target asphalt pavement image is obtained, the target core sample image is identified, and the identification result is used as the identification result of all target asphalt pavement images in a certain target asphalt pavement image set. In this way, while reducing the collection of core samples from the asphalt pavement, the water damage condition of the asphalt pavement can be accurately identified.

[0045] Please refer to Figure 2 , which shows the structural block diagram of an identification system for asphalt pavement water damage of the present application.

[0046] As Figure 2 shown, the identification system 200 for asphalt pavement water damage includes an acquisition module 210, a division module 220, a first judgment module 230, a second judgment module 240, an update module 250, and an identification module 260.

[0047] Among them, the acquisition module 210 is configured to acquire asphalt pavement images of a continuous section of asphalt pavement, input each asphalt pavement image into a preset asphalt image recognition model, and the asphalt image recognition model outputs the disease areas in each asphalt pavement image; the division module 220 is configured to divide each asphalt pavement image according to the positions of the disease areas in the asphalt pavement image by using a preset area division rule to obtain at least one asphalt pavement image set; the first judgment module 230 is configured to select a first asphalt pavement image and a second asphalt pavement image in a certain asphalt pavement image set, and judge whether the first asphalt pavement image and the second asphalt pavement image are the same asphalt pavement image, where the first asphalt pavement image is the asphalt pavement image with the largest number of disease areas, and the second asphalt pavement image is the asphalt pavement image with the smallest number of disease areas; the second judgment module 240 is configured to, if they are not the same asphalt pavement image, determine the association degree between the first target disease area and other disease areas in the first asphalt pavement image, and judge whether each association degree is greater than a first preset threshold, where the first target disease area is the disease area that divides the first asphalt pavement image into the certain asphalt pavement image set, and the other disease areas are the disease areas in the first asphalt pavement image except the target disease area; the update module 250 is configured to, if a certain association degree is not greater than the first preset threshold, remove the first asphalt pavement image from the certain asphalt pavement image set, and update the certain asphalt pavement image set after removing the first asphalt pavement image according to a preset update strategy to obtain at least one updated target asphalt pavement image set; the recognition module 260 is configured to select a certain target asphalt pavement image in a certain target asphalt pavement image set, obtain the target core sample image at the position of the second target disease area in the certain target asphalt pavement image, recognize the target core sample image, and use the recognition result as the recognition result of all target asphalt pavement images in the certain target asphalt pavement image set.

[0048] It should be understood that Figure 2 the modules described in Figure 1 correspond to the respective steps in the method described in Figure 2 Accordingly, the operations, features, and corresponding technical effects described above for the method also apply to Figure 2 the modules described herein and will not be repeated here.

[0049] In some other embodiments, the embodiments of the present invention further provide a computer-readable storage medium, on which a computer program is stored. When the program instructions are executed by a processor, the processor is caused to execute the asphalt pavement water damage recognition method in any of the above method embodiments; As an implementation manner, the computer-readable storage medium of the present invention stores computer-executable instructions, and the computer-executable instructions are set as follows: Obtain asphalt pavement images of a continuous section of asphalt pavement, and input each asphalt pavement image into a preset asphalt image recognition model, and the asphalt image recognition model outputs the disease areas in each asphalt pavement image; According to the positions of the disease areas in the asphalt pavement images, use a preset area division rule to divide each asphalt pavement image to obtain at least one asphalt pavement image set; Select a first asphalt pavement image and a second asphalt pavement image in a certain asphalt pavement image set, and determine whether the first asphalt pavement image and the second asphalt pavement image are the same asphalt pavement image, where the first asphalt pavement image is the asphalt pavement image with the largest number of disease areas, and the second asphalt pavement image is the asphalt pavement image with the smallest number of disease areas; If they are not the same asphalt pavement image, determine the association degree between the first target disease area and other disease areas in the first asphalt pavement image, and judge whether each association degree is greater than a first preset threshold, where the first target disease area is the disease area that divides the first asphalt pavement image into the certain asphalt pavement image set, and the other disease areas are the disease areas in the first asphalt pavement image excluding the target disease area; If a certain association degree is not greater than the first preset threshold, remove the first asphalt pavement image from the certain asphalt pavement image set, and update the certain asphalt pavement image set after removing the first asphalt pavement image according to a preset update strategy to obtain at least one updated target asphalt pavement image set; Select a certain target asphalt pavement image in a certain target asphalt pavement image set, obtain the target core sample image at the position of the second target disease area in the certain target asphalt pavement image, identify the target core sample image, and use the identification result as the identification result of all target asphalt pavement images in the certain target asphalt pavement image set.

[0050] A computer-readable storage medium may include a program storage area and a data storage area. Among them, the program storage area may store an operating system and application programs required for at least one function; the data storage area may store data created according to the use of the asphalt pavement water damage identification system, etc. In addition, the computer-readable storage medium may include high-speed random access memory, and may also include a memory, such as at least one magnetic disk storage device, a flash memory device, or other non-volatile solid-state storage devices. In some embodiments, the computer-readable storage medium may optionally include a memory remotely provided with respect to the processor, and these remote memories may be connected to the asphalt pavement water damage identification system through a network. Examples of the above network include but are not limited to the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0051] Figure 3 is a schematic structural diagram of an electronic device provided by an embodiment of the present invention, as Figure 3 shown, the device includes: a processor 310 and a memory 320. The electronic device may further include: an input device 330 and an output device 340. The processor 310, the memory 320, the input device 330, and the output device 340 may be connected through a bus or other means, Figure 3 taking the connection through the bus as an example. The memory 320 is the above-mentioned computer-readable storage medium. The processor 310 executes various functional applications and data processing of the server by running non-volatile software programs, instructions, and modules stored in the memory 320, that is, implements the asphalt pavement water damage identification method in the above method embodiment. The input device 330 may receive input digital or character information, and generate key signal inputs related to user settings and function controls of the asphalt pavement water damage identification system. The output device 340 may include a display device such as a display screen.

[0052] The above electronic device may execute the method provided by the embodiment of the present invention, and has corresponding functional modules and beneficial effects for executing the method. For technical details not described in detail in this embodiment, reference may be made to the method provided by the embodiment of the present invention.

[0053] As an implementation manner, the above electronic device is applied to an asphalt pavement water damage identification system and is used for a client, including: at least one processor; and a memory communicatively connected to the at least one processor; wherein, the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to: Obtain asphalt pavement images of a continuous section of asphalt pavement, input each asphalt pavement image into a preset asphalt image recognition model, and the asphalt image recognition model outputs the disease areas in each asphalt pavement image; According to the positions of each disease area in the asphalt pavement image, the above-mentioned asphalt pavement images are divided by using a preset area division rule to obtain at least one asphalt pavement image set; Select a first asphalt pavement image and a second asphalt pavement image from a certain asphalt pavement image set, and determine whether the first asphalt pavement image and the second asphalt pavement image are the same asphalt pavement image. Among them, the first asphalt pavement image is the asphalt pavement image with the largest number of disease areas, and the second asphalt pavement image is the asphalt pavement image with the smallest number of disease areas; If they are not the same asphalt pavement image, determine the correlation degree between the first target disease area and other disease areas in the first asphalt pavement image, and judge whether each correlation degree is greater than a first preset threshold. Among them, the first target disease area is the disease area that divides the first asphalt pavement image into the certain asphalt pavement image set, and the other disease areas are the disease areas in the first asphalt pavement image except the target disease area; If a certain correlation degree is not greater than the first preset threshold, remove the first asphalt pavement image from the certain asphalt pavement image set, and update the certain asphalt pavement image set after removing the first asphalt pavement image according to a preset update strategy to obtain at least one updated target asphalt pavement image set; Select a certain target asphalt pavement image from a certain target asphalt pavement image set, obtain the target core sample image at the position of the second target disease area in the certain target asphalt pavement image, identify the target core sample image, and use the identification result as the identification result of all target asphalt pavement images in the certain target asphalt pavement image set.

[0054] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course also by hardware. Based on such an understanding, the above technical solutions, in essence, or the part that contributes to the prior art can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods of each embodiment or some parts of the embodiments.

[0055] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A method for identifying water damage of asphalt pavement, characterized in that, Including: Obtain asphalt pavement images of a continuous section of asphalt pavement, and input each asphalt pavement image into a preset asphalt image recognition model. The asphalt image recognition model outputs the disease areas in each asphalt pavement image. According to the positions of the disease areas in the asphalt pavement images, use a preset area division rule to divide the asphalt pavement images, and obtain at least one asphalt pavement image set. Select a first asphalt pavement image and a second asphalt pavement image from a certain asphalt pavement image set, and determine whether the first asphalt pavement image and the second asphalt pavement image are the same asphalt pavement image. Among them, the first asphalt pavement image is the asphalt pavement image with the largest number of disease areas, and the second asphalt pavement image is the asphalt pavement image with the smallest number of disease areas. If they are not the same asphalt pavement image, determine the association degree between the first target disease area and other disease areas in the first asphalt pavement image, and judge whether each association degree is greater than a first preset threshold. Among them, the first target disease area is the disease area that divides the first asphalt pavement image into the certain asphalt pavement image set, and the other disease areas are the disease areas in the first asphalt pavement image except the target disease area. If a certain association degree is not greater than the first preset threshold, remove the first asphalt pavement image from the certain asphalt pavement image set, and update the certain asphalt pavement image set after removing the first asphalt pavement image according to a preset update strategy to obtain at least one updated target asphalt pavement image set. Select a certain target asphalt pavement image from a certain target asphalt pavement image set, obtain the target core sample image at the position of the second target disease area in the certain target asphalt pavement image, identify the target core sample image, and use the identification result as the identification result of all target asphalt pavement images in the certain target asphalt pavement image set.

2. The identification method of asphalt pavement water damage according to claim 1, characterized in that, The step of using a preset area division rule to divide the asphalt pavement images according to the positions of the disease areas in the asphalt pavement images to obtain at least one asphalt pavement image set includes: Stitch the asphalt pavement images in the order of acquisition time, and place the stitched asphalt pavement images in a pre-constructed two-dimensional coordinate system. Obtain the coordinate information of the regional center points in each disease area according to the two-dimensional coordinate system. According to each coordinate information, use the area division rule to divide each regional center point to obtain at least one regional center point set. Divide the asphalt pavement images corresponding to the regional center points in a certain regional center point set into the same asphalt pavement image set, that is, obtain the at least one asphalt pavement image set, where the certain regional center point set is any one of the at least one regional center point sets.

3. The identification method of asphalt pavement water damage according to claim 2, characterized in that, The area division rule is specifically: Set different windows, and divide all regional center points into different windows according to the ordinate values of the regional center points. Determine whether the difference between the first abscissa value of the center point of the first region and the second abscissa value of the center point of the second region within the same window is greater than a second preset threshold, where the center point of the first region and the center point of the second region are two adjacent center points of regions; If it is greater than the second preset threshold, do not divide the center point of the first region and the center point of the second region into the same set of center points of regions; If it is not greater than the second preset threshold, divide the center point of the first region and the center point of the second region into the same set of center points of regions, that is, obtain at least one set of regional centers.

4. The identification method of asphalt pavement water damage according to claim 3, characterized in that The setting of different windows includes: Obtain the coordinate information of the center points of each disease region in all asphalt pavement images; Cluster at least two center points of regions in the same asphalt pavement image whose distances are not greater than a preset distance threshold according to the coordinate information to obtain at least one target regional center; Select an asphalt pavement image with the largest number of target regional centers in each asphalt pavement image and define it as the target asphalt pavement image; Set the window according to each target regional center in the target asphalt pavement image, where the length of the window along the horizontal axis direction of the two-dimensional coordinate system is an infinite length, and the width of the window along the vertical axis direction of the two-dimensional coordinate system is twice the distance from the target regional center to the upper boundary or the lower boundary of the window.

5. The identification method of asphalt pavement water damage according to claim 1, characterized in that After determining whether the first asphalt pavement image and the second asphalt pavement image are the same asphalt pavement image, the method further includes: If they are the same asphalt pavement image, do not remove the first asphalt pavement image from the set of a certain asphalt pavement images, and obtain the target core sample image of the location of the first target disease region in the first asphalt pavement image; Input the target core sample image into a preset asphalt pavement water damage identification model, and the asphalt pavement water damage identification model outputs the water damage identification result of the target core sample image; Use the water damage identification result as the identification result of all target asphalt pavement images in the set of a certain asphalt pavement images.

6. The identification method of asphalt pavement water damage according to claim 1, characterized in that The determination of the degree of association between the target disease region and other disease regions in the first asphalt pavement image includes: Judge whether there is partial overlap between the target disease region and a certain disease region; If there is partial overlap, directly define the value of the degree of association between the target disease region and the certain disease region as the first preset threshold; If there is no partial overlap, determine the target center point of the target disease region and a certain center point of a certain disease region, and define the distance between the target center point and the certain center point as the degree of association between the target disease region and the certain disease region.

7. A method for identifying water damage to an asphalt pavement according to claim 1, characterized in that, After determining whether each degree of association is greater than the first preset threshold, the method further includes: If each degree of association is greater than the first preset threshold, do not remove the first asphalt pavement image from the set of a certain asphalt pavement images, and obtain the target core sample image of the location of the first target disease region in the first asphalt pavement image; Input the target core sample image into a preset asphalt pavement water damage identification model, and the asphalt pavement water damage identification model outputs a water damage identification result for the target core sample image; Use the water damage identification result as the identification result for all target asphalt pavement images in a certain asphalt pavement image set.

8. The identification method of asphalt pavement water damage according to claim 1, characterized in that, Updating a certain asphalt pavement image set after removing the first asphalt pavement image according to a preset update strategy to obtain at least one updated target asphalt pavement image set includes: Determine whether the distance between the third target disease area in the third asphalt pavement image and the fourth target disease area in the fourth asphalt pavement image in a certain asphalt pavement image set after removing the first asphalt pavement image is greater than a second preset threshold, where the third asphalt pavement image and the fourth asphalt pavement image are two adjacent asphalt pavement images, and the third target disease area is the disease area that divides the third asphalt pavement image into the certain asphalt pavement image set, and the fourth target disease area is the disease area that divides the third asphalt pavement image into the certain asphalt pavement image set; If the distance between the third target disease area and the fourth target disease area is greater than the second preset threshold, then divide the third asphalt pavement image and the fourth asphalt pavement image into different target asphalt pavement image sets respectively to obtain at least one target asphalt pavement image set; If the distance between the third target disease area and the fourth target disease area is not greater than the second preset threshold, then divide the third asphalt pavement image and the fourth asphalt pavement image into the same target asphalt pavement image set respectively to obtain at least one target asphalt pavement image set.

9. A method for identifying water damage to an asphalt pavement according to claim 1, characterized in that, Wherein, The second target disease area is the disease area that divides a certain target asphalt pavement image into the certain target asphalt pavement image set; Obtaining a target core sample image at the location of the target disease area in a certain target asphalt pavement image, identifying the target core sample image, and using the identification result as the identification result for all target asphalt pavement images in the certain target asphalt pavement image set includes: Obtain a target core sample image at the location of the target disease area in a certain target asphalt pavement image, where the certain target asphalt pavement image is any target asphalt pavement image in the certain target asphalt pavement image set; Input the target core sample image into a preset asphalt pavement water damage identification model, and the asphalt pavement water damage identification model outputs a water damage identification result for the target core sample image; Use the water damage identification result as the identification result for all target asphalt pavement images in the certain target asphalt pavement image set.

10. An identification system for water damage of asphalt pavement, characterized in that, Includes: An acquisition module configured to acquire asphalt pavement images of a continuous section of asphalt pavement, input each asphalt pavement image into a preset asphalt image recognition model, and the asphalt image recognition model outputs the disease areas in each asphalt pavement image; A partitioning module, configured to partition each of the asphalt pavement images according to a preset region partitioning rule based on the positions of the respective disease regions in the asphalt pavement images, to obtain at least one asphalt pavement image set; A first determination module, configured to select a first asphalt pavement image and a second asphalt pavement image in a certain asphalt pavement image set, and determine whether the first asphalt pavement image and the second asphalt pavement image are the same asphalt pavement image, wherein the first asphalt pavement image is the asphalt pavement image with the largest number of disease regions, and the second asphalt pavement image is the asphalt pavement image with the smallest number of disease regions; A second determination module, configured to, if they are not the same asphalt pavement image, determine the degree of association between the first target disease region and other disease regions in the first asphalt pavement image, and determine whether each degree of association is greater than a first preset threshold, wherein the first target disease region is the disease region that partitions the first asphalt pavement image into the certain asphalt pavement image set, and the other disease regions are the disease regions in the first asphalt pavement image excluding the target disease region; An update module, configured to, if a certain degree of association is not greater than the first preset threshold, remove the first asphalt pavement image from the certain asphalt pavement image set, and update the certain asphalt pavement image set after removing the first asphalt pavement image according to a preset update strategy, to obtain at least one updated target asphalt pavement image set; An identification module, configured to select a certain target asphalt pavement image in a certain target asphalt pavement image set, obtain a target core sample image at the position of the second target disease region in the certain target asphalt pavement image, identify the target core sample image, and use the identification result as the identification result of all target asphalt pavement images in the certain target asphalt pavement image set.

Images