Road surface damage detection method and system based on machine vision

Abstract

The invention discloses a road surface damage detection method and system based on machine vision. The method comprises the following steps: acquiring a target detection road surface photo corresponding to a target detection road surface area; judging whether a target suspected damaged area exists on the target detection pavement photo or not; if yes, obtaining a target detection road surface type of the target detection road surface area and a target suspected road surface damage type of the target suspected damage area; according to the target detection pavement type and the target suspected pavement damage type, comparing the target suspected damage area with a target damage area corresponding to a pre-stored target historical damage detection pavement photo, and judging whether the similarity between the target suspected damage area and the target damage area is greater than or equal to a second target preset photo similarity threshold; and if yes, determining the target suspected damage area as a target road surface damage area, and then determining the target detection road surface area as a road surface damage area. According to the invention, the damaged area of the road surface can be accurately detected and accurately positioned.

Description

Technical Field

[0001] The present invention relates to the technical field of road pavement detection, and in particular to a road damage detection method and system based on machine vision. Background Art

[0002] The machine vision system converts the captured target into an image signal through a machine vision product (i.e., an image capture device), transmits it to a dedicated image processing system, obtains the morphological information of the captured target, and converts it into a digital signal based on the pixel distribution, brightness, color and other information; the image system performs various operations on these signals to extract the characteristics of the target, and then controls the equipment action on site according to the results of the judgment. With the development of image processing technology and computer technology, machine vision can complete many visual tasks. With the rapid development of cities, there are more and more cars, which leads to an increasing traffic volume, which will cause more and more pressure on urban roads. Due to disrepair, repeated excavation caused by the addition of municipal pipelines, and other reasons, some roads have different degrees of damage, which will bring certain safety hazards to the normal driving of vehicles. The existing road detection method is based on machine vision. The road inspection vehicle takes road photos, and then manually analyzes the photos to extract the road damage data. However, the above method has the following problems: manually analyzing and detecting road surface photos requires the naked eye to identify tiny cracks on the road surface photos, which is labor-intensive and takes a long time to identify; in addition, manual identification of a large number of road surface photos is prone to misjudgment and cannot obtain correct damage detection results, which is not conducive to the development of subsequent work. Summary of the invention

[0003] In view of the above problems, the present invention proposes a road damage detection method and system based on machine vision to solve the following problems of the prior art: manually analyzing and detecting road surface photos requires the naked eye to identify tiny cracks on the road surface photos, which is labor-intensive and takes a long time to identify; in addition, manual identification of a large number of road surface photos is prone to misjudgment and cannot obtain correct damage detection results, which is not conducive to the subsequent work.

[0004] The first technical solution of the embodiment of the present invention is:

[0005] A road damage detection method based on machine vision, comprising: obtaining a target detection road surface photo corresponding to a target detection road surface area at a target preset time interval, wherein the target detection road surface geographic coordinate data corresponding to the target detection road surface area is pre-stored in a target database; judging whether there is a target suspected damaged area on the target detection road surface photo, wherein the similarity between the target suspected damaged area and its adjacent area is less than a first target preset photo similarity threshold; if so, obtaining a target detection road surface type corresponding to the target detection road surface area and a target suspected road surface damage type corresponding to the target suspected damaged area, wherein the target detection road surface type includes a target asphalt detection road surface and a target cement detection road surface, and the target suspected road surface damage type Including target transverse crack damage type, target vertical crack damage type, target oblique crack damage type and target depression damage type; according to the target detected pavement type and the target suspected pavement damage type, the target suspected damaged area and the target damaged area corresponding to the pre-stored target historical damaged detected pavement photo are compared to determine whether the similarity between the target suspected damaged area and the target damaged area is greater than or equal to a second target preset photo similarity threshold, wherein the target historical damaged detected pavement photo is a historical pavement photo with a damaged area corresponding to the target detected pavement type; if so, the target suspected damaged area is determined to be a target pavement damaged area, and then the target detected pavement area is determined to be a pavement damaged area.

[0006] The second technical solution of the embodiment of the present invention is:

[0007] A road damage detection system based on machine vision, comprising: a road photo acquisition module, used to acquire a target detection road photo corresponding to a target detection road area at a target preset time interval, wherein the target detection road geographic coordinate data corresponding to the target detection road area is pre-stored in a target database; a suspected area judgment module, used to judge whether there is a target suspected damaged area on the target detection road photo, wherein the similarity between the target suspected damaged area and its adjacent area is less than a first target preset photo similarity threshold; a damage type acquisition module, used to acquire the target detection road type corresponding to the target detection road area and the target suspected road damage type corresponding to the target suspected damaged area when there is a target suspected damaged area on the target detection road photo, wherein the target detection road type includes a target asphalt detection road surface and a target cement detection road surface, and the target suspected road damage type includes a target transverse crack damage type, target vertical crack damage type, target oblique crack damage type and target depression damage type; a similarity judgment module, used to compare the target suspected damage area with the target damaged area corresponding to the pre-stored target historical damaged detection road surface photo according to the target detected road surface type and the target suspected road surface damage type, and judge whether the similarity between the target suspected damage area and the target damaged area is greater than or equal to a second target preset photo similarity threshold, wherein the target historical damaged detection road surface photo is a historical road surface photo with a damaged area corresponding to the target detected road surface type; a damaged area judgment module, used to judge the target suspected damage area as a target road surface damaged area when the similarity between the target suspected damage area and the target damaged area is greater than or equal to the second target preset photo similarity threshold, and then judge the target detected road surface area as a road surface damaged area.

[0008] The third technical solution of the embodiment of the present invention is:

[0009] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor performs the following steps: obtaining a target detection road surface photo corresponding to a target detection road surface area at a target preset time interval, wherein the target detection road surface geographic coordinate data corresponding to the target detection road surface area is pre-stored in a target database; determining whether there is a target suspected damaged area on the target detection road surface photo, wherein the similarity between the target suspected damaged area and its adjacent area is less than a first target preset photo similarity threshold; if so, obtaining a target detection road surface type corresponding to the target detection road surface area and a target suspected road surface damage type corresponding to the target suspected damaged area, wherein the target detection road surface type comprises a target asphalt detection road surface and a A target cement detection road surface, wherein the target suspected road surface damage type includes a target transverse crack damage type, a target vertical crack damage type, a target oblique crack damage type and a target depression damage type; according to the target detection road surface type and the target suspected road surface damage type, the target suspected damaged area is compared with the target damaged area corresponding to the pre-stored target historical damaged detection road surface photo, and it is determined whether the similarity between the target suspected damaged area and the target damaged area is greater than or equal to a second target preset photo similarity threshold, wherein the target historical damaged detection road surface photo is a historical road surface photo with a damaged area corresponding to the target detection road surface type; if so, the target suspected damaged area is determined to be a target road surface damaged area, and then the target detection road surface area is determined to be a road surface damaged area.

[0010] The fourth technical solution of the embodiment of the present invention is:

[0011] A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the processor performs the following steps: obtaining a target detection road surface photo corresponding to a target detection road surface area at a target preset time interval, wherein the target detection road surface geographic coordinate data corresponding to the target detection road surface area is pre-stored in a target database; determining whether there is a target suspected damaged area on the target detection road surface photo, wherein the similarity between the target suspected damaged area and its adjacent area is less than a first target preset photo similarity threshold; if so, obtaining a target detection road surface type corresponding to the target detection road surface area and a target suspected road surface damage type corresponding to the target suspected damaged area, wherein the target detection road surface type includes a target asphalt detection road surface and a target cement detection road surface. The target suspected road surface damage type includes a target transverse crack damage type, a target vertical crack damage type, a target oblique crack damage type and a target depression damage type; according to the target detected road surface type and the target suspected road surface damage type, the target suspected damaged area is compared with the target damaged area corresponding to the pre-stored target historical damaged detected road surface photo, and it is determined whether the similarity between the target suspected damaged area and the target damaged area is greater than or equal to a second target preset photo similarity threshold, wherein the target historical damaged detected road surface photo is a historical road surface photo with a damaged area corresponding to the target detected road surface type; if so, the target suspected damaged area is determined to be a target road surface damaged area, and then the target detected road surface area is determined to be a road surface damaged area.

[0012] Compared with the prior art, the present invention has the following beneficial effects:

[0013] The present invention first obtains a target detection road surface photo corresponding to the target detection road surface area at each target preset time interval, and then determines whether there is a target suspected damaged area on the target detection road surface photo. If so, the target detection road surface type corresponding to the target detection road surface area and the target suspected road surface damage type corresponding to the target suspected damaged area are obtained. Finally, according to the target detection road surface type and the target suspected road surface damage type, the target suspected damaged area is compared with the target damaged area corresponding to the pre-stored target historical damaged detection road surface photo to determine whether the similarity between the target suspected damaged area and the target damaged area is greater than or equal to a second target preset photo similarity threshold. If so, the target suspected damaged area is determined as a target road surface damaged area, and then the target detection road surface area is determined as a road surface damaged area. The present invention can accurately detect the damaged area of ​​the road surface and perform accurate positioning. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0015] in:

[0016] Figure 1 It is an implementation flow chart of an implementation method of a road damage detection method based on machine vision in one embodiment;

[0017] Figure 2 It is a framework structure diagram of an implementation scheme of a road damage detection system based on machine vision in one embodiment;

[0018] Figure 3 It is a structural block diagram of an implementation scheme of a computer device in one embodiment. DETAILED DESCRIPTION

[0019] 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 only 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 ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0020] See also Figure 1 , combined with Figure 1 It can be seen that a road damage detection method based on machine vision according to an embodiment of the present invention includes the following steps:

[0021] Step S101: acquiring a target detection road surface photo corresponding to a target detection road surface area at a target preset time interval, wherein the target detection road surface geographic coordinate data corresponding to the target detection road surface area is pre-stored in a target database.

[0022] The target preset duration is a preset time length, which can be selected as a duration in milliseconds, and is not limited here. In addition, in order to accurately locate the location of the road damage, the target detection road surface geographic coordinate data corresponding to the target detection road surface area is pre-stored in the target database. Once the damage problem is detected in the target detection road surface area, the damaged area can be accurately located.

[0023] Step S102: determining whether there is a target suspected damaged area on the target detection road surface photo, wherein the similarity between the target suspected damaged area and its adjacent areas is less than a first target preset similarity threshold.

[0024] The target suspected damaged area is an area on the target detection road surface photo, and the appearance of the target suspected damaged area is quite different from other areas on the target detection road surface photo, while the appearance of each part of the normal undamaged road surface is not much different, that is, the similarity between the target suspected damaged area and its adjacent areas is less than the first target preset similarity threshold. The first target preset similarity threshold is an empirical threshold and is not limited here.

[0025] Step S103: If there is a target suspected damaged area on the target detection road surface photo, the target detection road surface type corresponding to the target detection road surface area and the target suspected road surface damage type corresponding to the target suspected damaged area are obtained, the target detection road surface type includes a target asphalt detection road surface and a target cement detection road surface, and the target suspected road surface damage type includes a target transverse crack damage type, a target vertical crack damage type, a target oblique crack damage type and a target depression damage type.

[0026] Among them, this step first determines the road surface type corresponding to the target detection road surface area, that is, the target detection road surface type, and then determines the suspected damage type corresponding to the target detection road surface area.

[0027] Among them, the target transverse crack damage type refers to the damage type in which the crack direction of the road surface is perpendicular to the road direction, the target vertical crack damage type refers to the damage type in which the crack direction of the road surface is parallel to the road direction, the target oblique crack damage type refers to the damage type in which the crack direction of the road surface forms an acute angle with the road direction, and the target depression damage type refers to the damage type in which the road surface is depressed by heavy pressure.

[0028] Step S104: According to the target detected road surface type and the target suspected road surface damage type, the target suspected damaged area is compared with the target damaged area corresponding to the pre-stored target historical damaged detected road surface photo, and it is determined whether the similarity between the target suspected damaged area and the target damaged area is greater than or equal to a second target preset similarity threshold, wherein the target historical damaged detected road surface photo is a historical road surface photo with a damaged area corresponding to the target detected road surface type.

[0029] The second target preset similarity threshold is an empirical threshold and is not limited here.

[0030] The target historical damaged road surface detection photo is a historical road surface photo with a damaged area corresponding to the target detected road surface type, that is, the detected road surface type corresponding to the target historical damaged road surface detection photo and the detected road surface type corresponding to the target detected road surface area are of the same type, for example, both are asphalt roads or both are cement roads. The appearance photos (including color and shape) of damaged parts of the same road surface type often have similar characteristics, that is, they have commonalities, so the similarity between the target suspected damaged area and the target damaged area can be used to determine whether the target suspected damaged area is a real damaged area.

[0031] Step S105: If the similarity between the target suspected damaged area and the target damaged area is greater than or equal to a second target preset similarity threshold, the target suspected damaged area is determined as a target road surface damaged area, and then the target detected road surface area is determined as a road surface damaged area.

[0032] In this embodiment, optionally, the determining whether there is a target suspected damaged area on the target detection road surface photo includes:

[0033] First, the target detection road surface photo is preprocessed to obtain a target detection road surface grayscale photo, and the geographic coordinate data of the target detection road surface grayscale photo corresponds to the geographic coordinate data of the target detection road surface.

[0034] Among them, in the target detection road surface grayscale photo, the target detection road surface grayscale photo can be divided into different areas according to different grayscale values. If the grayscale values ​​of two areas are greatly different, it means that the two areas are greatly different.

[0035] Second, determine whether there is a target abnormal grayscale area in the grayscale photograph of the target detection road surface, the grayscale value of the target abnormal grayscale area is different from the grayscale value of its adjacent area, and the grayscale difference between the target abnormal grayscale area and the adjacent area is greater than the target preset grayscale difference, wherein all grayscale values ​​in the target abnormal grayscale area are the same value, or the grayscale difference between the maximum grayscale value and the minimum grayscale value in the target abnormal grayscale area is less than the second target preset grayscale difference, and the second target preset grayscale difference is less than the first target preset grayscale difference.

[0036] The first target preset grayscale difference and the second target preset grayscale difference are both empirical thresholds, which are not limited here.

[0037] Among them, for example, there is an abnormal grayscale area in the middle of the target abnormal grayscale area with all grayscale values ​​being the same (or the grayscale difference between the maximum grayscale value and the minimum grayscale value is less than the second target preset grayscale difference). The grayscale value of this abnormal grayscale area is assumed to be the abnormal grayscale, and the grayscale values ​​of other areas adjacent to this abnormal grayscale area are different from this abnormal grayscale, and the grayscale difference is also large.

[0038] Third, if there is a target abnormal grayscale area in the target detection road grayscale photo, it is determined whether the target abnormal grayscale area is a target irregular shape grayscale area, wherein the edge contour of the target irregular shape grayscale area is an irregular shape.

[0039] Among them, the damage to the road surface is generally cracks or depressions. The grayscale value of the cracked or depressed area is usually different from the grayscale value of other undamaged areas, and the shape of the cracks is often irregular, that is, the edge contour of the target irregular shape grayscale area is an irregular shape.

[0040] Fourth, if the target abnormal grayscale region is a target irregular shape grayscale region, it is determined whether the target irregular shape grayscale region is a water accumulation region.

[0041] Among them, the waterlogged area looks different from other parts in the photo because there is water on the road surface, which may cause detection errors. Therefore, it is necessary to determine whether the target irregular shape grayscale area is a waterlogged area.

[0042] Fifth, if the target irregular shape grayscale region is not a water accumulation region, it is determined whether the target irregular shape grayscale region is an irregular region formed by accumulation of foreign matter on the road surface.

[0043] Among them, the accumulation of foreign objects on the road surface can also cause errors in road damage detection. The accumulated foreign objects on the road surface occupy a certain area of ​​the detected road surface, and their grayscale values ​​are often the same. Moreover, the grayscale value of the road surface area where foreign objects are accumulated is often different from the grayscale values ​​of other places on the detected road surface. Therefore, it is necessary to determine whether the target irregular shape grayscale area is an irregular area formed by the accumulation of foreign objects on the road surface.

[0044] Among them, whether the target irregular shape grayscale area is an irregular area formed by accumulation of foreign matter on the road surface can be determined by manual judgment or by photo comparison.

[0045] Sixth, if the target irregular shape grayscale area is not an irregular area formed by accumulation of foreign matter on the road surface, it is determined that there is a suspected damaged area of ​​the target on the target detection road surface photo.

[0046] In this embodiment, optionally, the determining whether the target irregular-shaped grayscale region is a water accumulation region includes:

[0047] First, determine whether the area corresponding to the target irregular shape grayscale region gradually decreases within a preset time period.

[0048] The preset time length can be selected as a few minutes or more than ten minutes, or other time lengths, which are not limited here. Because the accumulated water will evaporate over time, the accumulated water area will become smaller and smaller, so as long as the area corresponding to the target irregular shape grayscale area gradually decreases within the preset time length, it means that it is a accumulated water area.

[0049] Second, if the area corresponding to the target irregular shape grayscale region gradually decreases within a preset time period, it is determined that the target irregular shape grayscale region is a water accumulation area.

[0050] In this embodiment, optionally, comparing the target suspected damaged area with the target damaged area corresponding to the pre-stored target historical damaged detected road surface photo according to the target detected road surface type and the target suspected road surface damage type, and determining whether the similarity between the target suspected damaged area and the target damaged area is greater than or equal to a second target preset photo similarity threshold includes:

[0051] First, according to the target detected road surface type and the target suspected road surface damage type, the pre-stored target historical damaged detected road surface photo and the target damaged area corresponding to the target historical damaged detected road surface photo are obtained.

[0052] The road surface type corresponding to the pre-stored target historical damaged road surface photo is the target detected road surface type, and the road surface damage type corresponding to the pre-stored target historical damaged road surface photo is the target suspected road surface damage type. Under the same road surface type and road surface damage type, the target damaged area corresponding to the target historical damaged road surface photo and the target suspected damaged area corresponding to the target suspected damaged area are highly comparable.

[0053] Second, the target suspected damaged area color corresponding to the target suspected damaged area is compared with the target damaged area color corresponding to the target damaged area to determine whether the similarity between the target suspected damaged area color and the target damaged area color is greater than or equal to a target preset color similarity threshold.

[0054] The target preset color similarity threshold is an empirical threshold and is not limited here. The color of the target suspected damaged area and the color of the target damaged area are both the original photo colors rather than grayscale colors.

[0055] Third, if the similarity between the color of the target suspected damaged area and the color of the target damaged area is greater than or equal to the target preset color similarity threshold, it is determined that the similarity between the target suspected damaged area and the target damaged area is greater than or equal to the second target preset photo similarity threshold.

[0056] In this embodiment, optionally, comparing the target suspected damaged area with the target damaged area corresponding to the pre-stored target historical damaged detected road surface photo according to the target detected road surface type and the target suspected road surface damage type, and determining whether the similarity between the target suspected damaged area and the target damaged area is greater than or equal to a second target preset photo similarity threshold includes:

[0057] First, according to the target detected road surface type and the target suspected road surface damage type, the pre-stored target historical damaged detected road surface photo and the target damaged area corresponding to the target historical damaged detected road surface photo are obtained.

[0058] The road surface type corresponding to the pre-stored target historical damaged road surface photo is the target detected road surface type, and the road surface damage type corresponding to the pre-stored target historical damaged road surface photo is the target suspected road surface damage type. Under the same road surface type and road surface damage type, the target damaged area corresponding to the target historical damaged road surface photo and the target suspected damaged area corresponding to the target suspected damaged area are highly comparable.

[0059] Second, compare the target suspected damaged area grayscale value corresponding to the target suspected damaged area with the target damaged area grayscale value corresponding to the target damaged area to determine whether the target difference absolute value between the target suspected damaged area grayscale value and the target damaged area grayscale value is less than the target preset difference absolute value.

[0060] The target preset difference absolute value is an empirical threshold value and is not limited here. When the target difference absolute value between the grayscale value of the target suspected damaged area and the grayscale value of the target damaged area is less than the target preset difference absolute value, it means that the grayscale value of the target suspected damaged area and the grayscale value of the target damaged area are substantially the same, and it can be determined that the similarity between the target suspected damaged area and the target damaged area is greater than or equal to the second target preset photo similarity threshold value.

[0061] Third, if the target absolute value of the difference between the grayscale value of the target suspected damaged area and the grayscale value of the target damaged area is less than the target preset absolute value of the difference, it is determined that the similarity between the target suspected damaged area and the target damaged area is greater than or equal to the second target preset photo similarity threshold.

[0062] In this embodiment, optionally, the determining whether there is a target suspected damaged area on the target detection road surface photo includes:

[0063] First, a historical photo of the target road surface under normal detection is obtained, where the historical photo of the target road surface under normal detection is a non-damaged historical photo corresponding to the target road surface area under detection.

[0064] Among them, the shooting environment brightness of the target historical normal detection road surface photo and the shooting environment brightness of the target detection road surface photo are the same or have a small difference, that is, the target historical normal detection road surface photo and the target detection road surface photo are both taken under the same or similar light brightness.

[0065] Second, the target detection road surface photo is compared with the target historical normal detection road surface photo to determine whether the similarity between the target detection road surface photo and the target historical normal detection road surface photo is less than a third target preset similarity threshold.

[0066] The third target preset similarity threshold is an empirical threshold and is not limited here.

[0067] Third, if the similarity between the target detection road surface photo and the target historical normal detection road surface photo is less than a third target preset similarity threshold, it is determined that there is a suspected damaged area of ​​the target on the target detection road surface photo.

[0068] In this embodiment, optionally, the determining whether there is a target suspected damaged area on the target detection road surface photo includes:

[0069] First, a target historical damaged road surface detection photo is obtained, where the target historical damaged road surface detection photo is a damaged historical photo corresponding to the target detection road surface area.

[0070] Among them, the shooting environment brightness of the target historical damage detection road surface photo and the shooting environment brightness of the target detection road surface photo are the same or have a small difference, that is, the target historical damage detection road surface photo and the target detection road surface photo are both taken under the same or similar light brightness.

[0071] Second, the target detection road surface photo is compared with the target historical damage detection road surface photo to determine whether the similarity between the target detection road surface photo and the target historical damage detection road surface photo is greater than or equal to a fourth target preset similarity threshold.

[0072] The fourth target preset similarity threshold is an empirical threshold and is not limited here.

[0073] Third, if the similarity between the target detection road surface photo and the target historical damaged detection road surface photo is greater than or equal to a fourth target preset similarity threshold, it is determined that there is a suspected damaged area of ​​the target on the target detection road surface photo.

[0074] See also Figure 2 , combined with Figure 2 It can be seen that a road damage detection system 100 based on machine vision according to an embodiment of the present invention includes:

[0075] A road surface photo acquisition module 10 is used to acquire a target detection road surface photo corresponding to the target detection road surface area at a target preset time interval, wherein the target detection road surface geographic coordinate data corresponding to the target detection road surface area is pre-stored in the target database;

[0076] A suspected area determination module 20 is used to determine whether there is a target suspected damaged area on the target detection road surface photo, wherein the similarity between the target suspected damaged area and its adjacent area is less than a first target preset photo similarity threshold;

[0077] A damage type acquisition module 30 is used to acquire the target detection road surface type corresponding to the target detection road surface area and the target suspected road surface damage type corresponding to the target suspected damage area when there is a target suspected damage area on the target detection road surface photo, wherein the target detection road surface type includes a target asphalt detection road surface and a target cement detection road surface, and the target suspected road surface damage type includes a target transverse crack damage type, a target vertical crack damage type, a target oblique crack damage type and a target depression damage type;

[0078] A similarity judgment module 40 is used to compare the target suspected damaged area with the target damaged area corresponding to the pre-stored target historical damaged detected road surface photo according to the target detected road surface type and the target suspected road surface damage type, and judge whether the similarity between the target suspected damaged area and the target damaged area is greater than or equal to a second target preset photo similarity threshold, wherein the target historical damaged detected road surface photo is a historical road surface photo with a damaged area corresponding to the target detected road surface type;

[0079] The damaged area determination module 50 is used to determine the target suspected damaged area as a target road surface damaged area when the similarity between the target suspected damaged area and the target damaged area is greater than or equal to the second target preset photo similarity threshold, and then determine the target detected road surface area as a road surface damaged area.

[0080] Figure 3 FIG. 1 shows an internal structure diagram of a computer device in an embodiment. The computer device may be a terminal or a server. Figure 3As shown, the computer device includes a processor, a memory and a network interface connected via a system bus. The memory includes a non-volatile storage medium and an internal memory. The non-volatile storage medium of the computer device stores an operating system and may also store a computer program. When the computer program is executed by the processor, the processor may implement the above-mentioned road damage detection method based on machine vision. The internal memory may also store a computer program. When the computer program is executed by the processor, the processor may implement the above-mentioned road damage detection method based on machine vision. Those skilled in the art will understand that Figure 3 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0081] In another embodiment, a computer device is proposed, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor performs the following steps: obtaining a target detection road surface photo corresponding to a target detection road surface area at a target preset time interval, wherein the target detection road surface geographic coordinate data corresponding to the target detection road surface area is pre-stored in a target database; determining whether there is a target suspected damaged area on the target detection road surface photo, wherein the similarity between the target suspected damaged area and its adjacent area is less than a first target preset photo similarity threshold; if so, obtaining a target detection road surface type corresponding to the target detection road surface area and a target suspected road surface damage type corresponding to the target suspected damaged area, wherein the target detection road surface type includes a target asphalt The target suspected pavement damage type includes a target transverse crack damage type, a target vertical crack damage type, a target oblique crack damage type and a target depression damage type; according to the target detection pavement type and the target suspected pavement damage type, the target suspected damaged area is compared with the target damaged area corresponding to the pre-stored target historical damaged detection pavement photo, and it is determined whether the similarity between the target suspected damaged area and the target damaged area is greater than or equal to a second target preset photo similarity threshold, wherein the target historical damaged detection pavement photo is a historical pavement photo with a damaged area corresponding to the target detection pavement type; if so, the target suspected damaged area is determined to be a target pavement damaged area, and then the target detection pavement area is determined to be a pavement damaged area.

[0082] In another embodiment, a computer-readable storage medium is proposed, storing a computer program, and when the computer program is executed by a processor, the processor performs the following steps: obtaining a target detection road surface photo corresponding to a target detection road surface area at a target preset time interval, wherein the target detection road surface geographic coordinate data corresponding to the target detection road surface area is pre-stored in a target database; determining whether there is a target suspected damaged area on the target detection road surface photo, wherein the similarity between the target suspected damaged area and its adjacent area is less than a first target preset photo similarity threshold; if so, obtaining the target detection road surface type corresponding to the target detection road surface area and the target suspected road surface damage type corresponding to the target suspected damaged area, wherein the target detection road surface type includes a target asphalt detection road surface and a target suspected damaged road surface. A target cement detection road surface, wherein the target suspected road surface damage type includes a target transverse crack damage type, a target vertical crack damage type, a target oblique crack damage type and a target depression damage type; according to the target detection road surface type and the target suspected road surface damage type, the target suspected damaged area is compared with the target damaged area corresponding to the pre-stored target historical damaged detection road surface photo, and it is determined whether the similarity between the target suspected damaged area and the target damaged area is greater than or equal to a second target preset photo similarity threshold, wherein the target historical damaged detection road surface photo is a historical road surface photo with a damaged area corresponding to the target detection road surface type; if so, the target suspected damaged area is determined to be a target road surface damaged area, and then the target detection road surface area is determined to be a road surface damaged area.

[0083] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing related hardware through a computer program, and the program can be stored in a non-volatile computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), Synchlink DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct RAM bus dynamic RAM (DRDRAM), and RAM bus dynamic RAM (RDRAM), etc.

[0084] The embodiment of the present invention first obtains a target detection road surface photo corresponding to the target detection road surface area at every target preset time period, and then determines whether there is a target suspected damaged area on the target detection road surface photo. If so, the target detection road surface type corresponding to the target detection road surface area and the target suspected road surface damage type corresponding to the target suspected damaged area are obtained. Finally, according to the target detection road surface type and the target suspected road surface damage type, the target suspected damaged area is compared with the target damaged area corresponding to the pre-stored target historical damaged detection road surface photo to determine whether the similarity between the target suspected damaged area and the target damaged area is greater than or equal to a second target preset photo similarity threshold. If so, the target suspected damaged area is determined as a target road surface damaged area, and then the target detection road surface area is determined as a road surface damaged area. The present invention can accurately detect the damaged area of ​​the road surface and perform accurate positioning.

[0085] The technical features of the above embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0086] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A road damage detection method based on machine vision, characterized in that: include: At each target preset time period, a target detection road surface photo corresponding to the target detection road surface area is obtained, wherein the target detection road surface geographic coordinate data corresponding to the target detection road surface area is pre-stored in the target database; Determine whether there is a target suspected damaged area on the target detection road surface photo, wherein the similarity between the target suspected damaged area and its adjacent area is less than a first target preset similarity threshold; If so, then obtain the target detected pavement type corresponding to the target detected pavement area and the target suspected pavement damage type corresponding to the target suspected damaged area, the target detected pavement type includes a target asphalt detected pavement and a target cement detected pavement, and the target suspected pavement damage type includes a target transverse crack damage type, a target vertical crack damage type, a target oblique crack damage type and a target depression damage type; According to the target detected road surface type and the target suspected road surface damage type, the target suspected damaged area is compared with the target damaged area corresponding to the pre-stored target historical damaged detected road surface photo, and it is determined whether the similarity between the target suspected damaged area and the target damaged area is greater than or equal to a second target preset similarity threshold, wherein the target historical damaged detected road surface photo is a historical road surface photo with a damaged area corresponding to the target detected road surface type; If so, the target suspected damaged area is determined as a target road surface damaged area, and then the target detected road surface area is determined as a road surface damaged area.

2. The machine vision-based road damage detection method according to claim 1, characterized in that: The determining whether there is a target suspected damaged area on the target detection road surface photo includes: Preprocessing the target detection road surface photo to obtain a target detection road surface grayscale photo, wherein the geographic coordinate data of the target detection road surface grayscale photo corresponds to the geographic coordinate data of the target detection road surface; Determine whether there is a target abnormal grayscale area in the target detection road grayscale photo, the grayscale value of the target abnormal grayscale area is different from the grayscale value of its adjacent area, and the grayscale difference between the target abnormal grayscale area and the adjacent area is greater than a first target preset grayscale difference, wherein all grayscale values ​​in the target abnormal grayscale area are the same value, or the grayscale difference between the maximum grayscale value and the minimum grayscale value in the target abnormal grayscale area is less than a second target preset grayscale difference, and the second target preset grayscale difference is less than the first target preset grayscale difference; If yes, determining whether the target abnormal grayscale region is a target irregular shape grayscale region, wherein the edge contour of the target irregular shape grayscale region is an irregular shape; If yes, determining whether the target irregular shape grayscale area is a water accumulation area; If not, determining whether the target irregular shape grayscale area is an irregular area formed by accumulation of foreign matter on the road surface; If not, it is determined that there is a suspected damaged area of ​​the target on the target detection road surface photo.

3. The machine vision-based road damage detection method according to claim 2, characterized in that: The step of judging whether the target irregular shape grayscale area is a water accumulation area includes: Determine whether the area corresponding to the target irregular shape grayscale region gradually decreases within a preset time length; If so, the target irregular shape grayscale area is determined to be a water accumulation area.

4. The machine vision-based road damage detection method according to claim 1, characterized in that: The method of comparing the target suspected damaged area with the target damaged area corresponding to the pre-stored target historical damaged detected road surface photo according to the target detected road surface type and the target suspected road surface damage type, and determining whether the similarity between the target suspected damaged area and the target damaged area is greater than or equal to a second target preset photo similarity threshold, includes: According to the target detected road surface type and the target suspected road surface damage type, obtaining a pre-stored target historical damaged detected road surface photo and the target damaged area corresponding to the target historical damaged detected road surface photo; Compare the target suspected damaged area color corresponding to the target suspected damaged area with the target damaged area color corresponding to the target damaged area, and determine whether the similarity between the target suspected damaged area color and the target damaged area color is greater than or equal to a target preset color similarity threshold; If so, it is determined that the similarity between the target suspected damaged area and the target damaged area is greater than or equal to the second target preset photo similarity threshold.

5. The machine vision-based road damage detection method according to claim 1, characterized in that: The method of comparing the target suspected damaged area with the target damaged area corresponding to the pre-stored target historical damaged detected road surface photo according to the target detected road surface type and the target suspected road surface damage type, and determining whether the similarity between the target suspected damaged area and the target damaged area is greater than or equal to a second target preset photo similarity threshold, includes: According to the target detected road surface type and the target suspected road surface damage type, obtaining a pre-stored target historical damaged detected road surface photo and the target damaged area corresponding to the target historical damaged detected road surface photo; Compare the target suspected damaged area grayscale value corresponding to the target suspected damaged area with the target damaged area grayscale value corresponding to the target damaged area, and determine whether the target difference absolute value between the target suspected damaged area grayscale value and the target damaged area grayscale value is less than the target preset difference absolute value; If so, it is determined that the similarity between the target suspected damaged area and the target damaged area is greater than or equal to the second target preset photo similarity threshold.

6. The machine vision-based road damage detection method according to claim 1, characterized in that: The determining whether there is a target suspected damaged area on the target detection road surface photo includes: Acquire a target historical normal detection road surface photo, wherein the target historical normal detection road surface photo is a non-damaged historical photo corresponding to the target detection road surface area; Comparing the target detection road surface photo with the target historical normal detection road surface photo, and determining whether the similarity between the target detection road surface photo and the target historical normal detection road surface photo is less than a third target preset similarity threshold; If so, it is determined that there is a suspected damaged area of ​​the target on the target detection road surface photo.

7. The machine vision-based road damage detection method according to claim 1, characterized in that: The determining whether there is a target suspected damaged area on the target detection road surface photo includes: Acquire a target historical damaged road surface detection photo, where the target historical damaged road surface detection photo is a damaged historical photo corresponding to the target detection road surface area; Comparing the target detection road surface photo with the target historical damage detection road surface photo, and determining whether the similarity between the target detection road surface photo and the target historical damage detection road surface photo is greater than or equal to a fourth target preset similarity threshold; If so, it is determined that there is a suspected damaged area of ​​the target on the target detection road surface photo.

8. A road damage detection system based on machine vision, characterized in that: include: A road surface photo acquisition module, used for acquiring a target detection road surface photo corresponding to the target detection road surface area at a target preset time interval, wherein the target detection road surface geographic coordinate data corresponding to the target detection road surface area is pre-stored in the target database; A suspected area judgment module is used to judge whether there is a target suspected damaged area on the target detection road surface photo, wherein the similarity between the target suspected damaged area and its adjacent area is less than a first target preset photo similarity threshold; A damage type acquisition module is used to acquire the target detection road surface type corresponding to the target detection road surface area and the target suspected road surface damage type corresponding to the target suspected damage area when there is a target suspected damage area on the target detection road surface photo, wherein the target detection road surface type includes a target asphalt detection road surface and a target cement detection road surface, and the target suspected road surface damage type includes a target transverse crack damage type, a target vertical crack damage type, a target oblique crack damage type and a target depression damage type; a similarity judgment module, for comparing the target suspected damaged area with the target damaged area corresponding to a pre-stored target historical damaged detected road surface photo according to the target detected road surface type and the target suspected road surface damage type, and judging whether the similarity between the target suspected damaged area and the target damaged area is greater than or equal to a second target preset photo similarity threshold, wherein the target historical damaged detected road surface photo is a historical road surface photo with a damaged area corresponding to the target detected road surface type; The damaged area determination module is used to determine the target suspected damaged area as a target road surface damaged area when the similarity between the target suspected damaged area and the target damaged area is greater than or equal to the second target preset photo similarity threshold, and then determine the target detected road surface area as a road surface damaged area.

9. A computer-readable storage medium, characterized in that: A computer program is stored, and when the computer program is executed by a processor, the processor executes the road damage detection method based on machine vision as described in any one of claims 1 to 7.

10. A computer device, characterized in that: It comprises a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the road damage detection method based on machine vision as described in any one of claims 1 to 7.

Images