Method and system for detecting quality of highway pavement based on machine vision

Abstract

The present application relates to the technical field of multi-image analysis, in particular to a highway pavement quality detection method and system based on machine vision. According to the present application, the brightness crack probability is obtained according to the gray scale difference of the same suspected crack area in the pavement gray scale images before and after watering, and the temperature crack probability is obtained according to the pixel value difference of the same suspected crack area in the pavement infrared images before and after watering and the environmental data; according to the significant degree of the pixel value difference of the same suspected crack area in the pavement feature images before and after watering, the brightness crack probability and the temperature crack probability are adjusted to obtain the final crack probability of the suspected crack area, and the pavement quality of the to-be-tested highway is detected by using the same. The present application combines the brightness change characteristics and temperature change characteristics of the crack before and after watering, reduces the dependence on light, and improves the accuracy of pavement crack detection.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of multi-image analysis, and in particular to a highway pavement quality detection method and system based on machine vision. BACKGROUND

[0002] Pavement cracks can cause steel corrosion, base softening, and ultimately lead to pavement collapse or bearing failure. Highway pavement quality detection is a core technical means to ensure road safety, durability and economic operation. Existing methods detect pavement cracks from visible light images of the pavement based on crack trends, width and other characteristics. However, in low light conditions and complex texture structure pavement, insufficient light and pavement texture structure can easily lead to low accuracy of crack feature recognition in visible light images, thereby reducing the accuracy of highway pavement quality detection. SUMMARY

[0003] To solve the technical problem of low accuracy of crack detection in visible light images due to insufficient light and pavement texture structure, the present application aims to provide a highway pavement quality detection method and system based on machine vision. The technical solution adopted is as follows:

[0004] In a first aspect, an embodiment of the present application provides a highway pavement quality detection method based on machine vision, which comprises:

[0005] Obtaining pavement feature images of the to-be-detected highway before and after watering, and environmental data of the to-be-detected highway location; the pavement feature images include pavement grayscale images and pavement infrared images;

[0006] Obtaining suspected crack areas in the pavement feature images; obtaining the brightness crack probability of each suspected crack area according to the grayscale difference of the same suspected crack area in the pavement grayscale images before and after watering;

[0007] Obtaining the temperature crack probability of each suspected crack area according to the pixel value difference of the same suspected crack area in the pavement infrared images before and after watering and the environmental data;

[0008] Adjusting the brightness crack probability and the temperature crack probability according to the degree of significance of the pixel value difference of the same suspected crack area in the pavement feature images before and after watering, to obtain the final crack probability of each suspected crack area; and using the final crack probability to detect the pavement quality of the to-be-detected highway.

[0009] Further, the obtaining of the brightness crack probability of each suspected crack area comprises:

[0010] Taking the mean value of the pixel values of all pixel points in the suspected crack area as the pixel overall value;

[0011] An absolute value of a difference between the pixel overall values of the same suspected crack region in the road surface feature images before and after watering is calculated to obtain a pixel difference of each suspected crack region in the road surface feature images;

[0012] The pixel difference of each suspected crack region in the road surface grayscale image is normalized to obtain a brightness crack probability of each suspected crack region.

[0013] Further, the temperature crack probability of each suspected crack region is obtained by:

[0014] The environmental data includes temperature data and wind speed data; and an environmental influence value is obtained according to the temperature data and the wind speed data.

[0015] The pixel difference of each suspected crack region in the road surface infrared image is adjusted by using the environmental influence value to obtain an actual temperature difference of each suspected crack region; and the actual temperature difference is normalized to obtain a temperature crack probability of each suspected crack region.

[0016] Further, the final crack probability of each suspected crack region is obtained by:

[0017] A feature weight coefficient is obtained according to a significant degree of a difference between pixel values of the same suspected crack region in the road surface feature images before and after watering.

[0018] The feature weight coefficient, a constant 1 and a difference between the feature weight coefficient are sequentially taken as a weight of the temperature crack probability and the brightness crack probability of each suspected crack region, respectively; the temperature crack probability and the brightness crack probability of each suspected crack region are weighted and summed to obtain a final crack probability of each suspected crack region.

[0019] Further, the method for obtaining the feature weight coefficient comprises:

[0020] A ratio of a mean value of the pixel differences of all suspected crack regions in the road surface feature images, a standard deviation and a sum of a preset positive number is taken as a feature significant degree; and the corresponding feature significant degrees of the road surface grayscale image and the road surface infrared image are sequentially taken as a grayscale feature significant degree and an infrared feature significant degree, respectively.

[0021] A ratio of the infrared feature significant degree as a numerator and a sum of the infrared feature significant degree and the grayscale feature significant degree as a denominator is taken as the feature weight coefficient.

[0022] Further, the suspected crack region of the road surface feature image is obtained by:

[0023] The maximum inter-class variance method is used to obtain a segmentation threshold value for the gray value of a pixel point in the road gray image before watering, and a connected domain formed by the pixel points with a gray value less than the segmentation threshold value is regarded as a suspected crack region of the road gray image before watering.

[0024] The road infrared images before and after watering and the road gray image after watering are denoted as analysis images, and a connected domain formed by the pixel points at the same coordinate positions in each suspected crack region of the road gray image before watering in the analysis images is regarded as a suspected crack region of the analysis image.

[0025] Further, the actual temperature difference of each suspected crack region is obtained by:

[0026] The environmental influence value is negatively correlated and normalized, and the pixel difference of each suspected crack region in the road infrared image is weighted by using the processing result, so that the actual temperature difference of each suspected crack region is obtained.

[0027] Further, the temperature data and the wind speed data are positively correlated with the environmental influence value.

[0028] Further, the pixel points in each two road feature images are one-to-one corresponding.

[0029] In the second aspect, another embodiment of the present application provides a highway pavement quality detection system based on machine vision, which comprises:

[0030] A data acquisition module is configured to acquire road feature images of a to-be-detected highway before and after watering and environmental data of the to-be-detected highway position, wherein the road feature images comprise road gray images and road infrared images;

[0031] A brightness feature analysis module is configured to acquire suspected crack regions in the road feature images, and to acquire a brightness crack probability of each suspected crack region according to the gray value difference of the same suspected crack region in the road gray images before and after watering;

[0032] A temperature feature analysis module is configured to acquire a temperature crack probability of each suspected crack region according to the pixel value difference of the same suspected crack region in the road infrared images before and after watering and the environmental data;

[0033] A pavement quality detection module is configured to adjust the brightness crack probability and the temperature crack probability according to the significant degree of the pixel value difference of the same suspected crack region in the road feature images before and after watering, to acquire a final crack probability of each suspected crack region, and to perform pavement quality detection on the to-be-detected highway by using the final crack probability.

[0034] The present application has the following beneficial effects:

[0035] The first aspect: compared with the traditional method of detecting cracks from the visible light image of the road surface according to the crack trend, width and other characteristics, the present scheme is based on the significant difference between the brightness change and the temperature change of the road surface and the crack after watering, the brightness change characteristics and the temperature change characteristics are used jointly, and the accuracy of the road crack detection is improved through multi-dimensional data cross verification.

[0036] The second aspect: the present scheme combines brightness change and temperature change for crack detection, compared with the traditional method of directly detecting cracks according to the visible light image, the present scheme additionally considers temperature change, and the dependence on light conditions is lower, and the road crack characteristics can still be accurately captured under low light conditions and complex road texture structure, thereby effectively avoiding the occurrence of missed detection, and the accuracy of the road crack detection is improved.

[0037] The third aspect: the present scheme uses thermal imaging technology to analyze the temperature change before and after watering at the same position, the thermal imaging technology can penetrate surface pollutants such as dust and oil stains, directly reflect the thermal characteristic change of the road crack caused by water evaporation, and can obtain more real road condition information, thereby improving the accuracy of the road crack detection.

[0038] The fourth aspect: when analyzing the thermal characteristic change of the road crack caused by water evaporation, the influence of environmental factors on the thermal characteristic change caused by water evaporation is considered, thereby improving the authenticity of the road state. BRIEF DESCRIPTION OF DRAWINGS

[0039] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiment or prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor on the basis of these drawings.

[0040] Figure 1 A step flow chart of a highway pavement quality detection method based on machine vision provided by an embodiment of the present application;

[0041] Figure 2 A flow chart of a final crack probability acquisition method provided by an embodiment of the present application;

[0042] Figure 3 A system structure diagram of a highway pavement quality detection system based on machine vision provided by an embodiment of the present application;

[0043] Figure 4 A computer device schematic diagram of a highway pavement quality detection device based on machine vision provided by an embodiment of the present application. DETAILED DESCRIPTION

[0044] In order to further illustrate the technical means and effects taken by the present application to achieve the predetermined inventive objectives, the specific implementation, structure, features and effects of a highway pavement quality detection method and system based on machine vision according to the present application are described in detail as follows in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures or characteristics in one or more embodiments can be combined in any suitable form.

[0045] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present application belongs.

[0046] The specific scenario to which the present application is directed: suitable for pavement quality detection of traffic infrastructure such as expressway, urban trunk road, bridge, tunnel, etc., especially for rapid detection of pavement damage such as pavement aging, crack propagation, etc. caused by long-term exposure to complex environment such as large temperature difference, high traffic flow, frequent rain and snow erosion, etc.

[0047] The specific scheme of a highway pavement quality detection method and system based on machine vision provided by the present application is described in detail below in combination with the accompanying drawings.

[0048] Embodiment 1

[0049] The present application proposes a highway pavement quality detection method based on machine vision. Please refer to Figure 1 which shows a step flowchart of a highway pavement quality detection method based on machine vision provided by one embodiment of the present application, which method comprises:

[0050] Step S1: respectively acquiring pavement feature images before and after watering of the to-be-detected highway, and environmental data of the to-be-detected highway position; the pavement feature images include pavement grayscale images and pavement infrared images.

[0051] An industrial camera and an infrared camera are integrated on the unmanned aerial vehicle, the unmanned aerial vehicle is controlled to fly above the to-be-detected highway and the position of the unmanned aerial vehicle is kept unchanged, and before a road sweeping and watering vehicle passes, the industrial camera and the infrared camera are used simultaneously to collect pavement images and pavement infrared images in turn; the pavement is sprayed with water during the passing of the road sweeping and watering vehicle, and after the road sweeping and watering vehicle passes for one minute, the industrial camera and the infrared camera are used simultaneously again to collect pavement images and pavement infrared images. The pavement images before and after watering are subjected to grayscale and denoising processing to obtain corresponding pavement grayscale images, and the pavement infrared images before and after watering are subjected to denoising processing, and the pavement grayscale images and the pavement infrared images are collectively referred to as pavement feature images. The one minute can also be two minutes, which can be set by the implementer according to the specific circumstances, and the pavement feature images after watering need to be collected before the water on the pavement evaporates completely.

[0052] In the time period from the moment when the water spraying vehicle finishes spraying water on the road surface of the to-be-tested road to the moment when the road surface feature image after spraying water is collected, the handheld portable weather station is used to collect the ambient temperature and ambient wind speed of the to-be-tested road at each time, and the average ambient temperature and the average ambient wind speed at all times in the time period are sequentially taken as the temperature data and the wind speed data, collectively referred to as the environmental data. The data collection frequency of the handheld portable weather station is set to once every 10 seconds.

[0053] It should be noted that the position of the unmanned aerial vehicle when collecting the road surface feature images before and after spraying water is the same, so as to ensure that the pixel points in the road surface feature images before and after spraying water correspond to each other, and at the same time, it is necessary to ensure that the pixel points in the road surface gray-scale images and the road surface infrared images before and after spraying water correspond to each other, so that the pixel points in each two road surface feature images correspond to each other. In the embodiment of the present application, a weighted average gray-scale algorithm is used for gray-scale processing, and a Gaussian filter is used for denoising. The specific method is not introduced here, and it is a well-known technical means to those skilled in the art.

[0054] Step S2: Obtain a suspected crack area in the road surface feature image; obtain a brightness crack probability of each suspected crack area according to the gray-scale difference of the same suspected crack area in the road surface gray-scale images before and after spraying water.

[0055] Although the road crack can be shown in the visible light image, due to the complexity of the road texture structure, the visible light image is difficult to accurately judge the crack, and the suspected crack area can be preliminarily extracted from the road surface feature image to reduce the subsequent calculation amount. Before the road is sprayed, the brightness difference between the crack and the road surface in the visible light image can be small; but after the road is sprayed, the water remaining in the crack forms a relatively closed environment, the propagation path of light in the crack is lengthened, the absorption and scattering are increased, resulting in a decrease in the brightness of the crack, while the brightness of the road surface remains basically unchanged due to rapid drainage and rapid evaporation of water. Therefore, the brightness difference of the crack compared to the road surface before and after spraying water is larger. According to the gray-scale difference of the same suspected crack area in the road surface gray-scale images before and after spraying water, the possibility of the suspected crack area being an actual crack is analyzed through the brightness change of the suspected crack area before and after spraying water, and a brightness crack probability is obtained.

[0056] Step S3: Obtain a temperature crack probability of each suspected crack area according to the pixel value difference of the same suspected crack area in the road surface infrared images before and after spraying water and the environmental data.

[0057] After the road surface is watered, the water quickly penetrates into the inside of the crack and stays, while the water on the road surface quickly flows away due to gravity, and the evaporation process of the water in the crack absorbs a large amount of heat, while the evaporation process of the small amount of water remaining on the road surface absorbs less heat, so that the temperature difference between the crack and the road surface before and after watering is larger. Considering that the external environmental changes such as the environmental temperature and the wind speed will affect the water evaporation process, resulting in an error in the temperature difference before and after watering at the same position, which affects the accuracy of crack judgment. Therefore, the pixel value difference of the same suspected crack area in the road surface infrared images before and after watering is analyzed to determine the possibility of the suspected crack area being an actual crack by using the temperature change of the suspected crack area before and after watering, and the environmental data is adjusted to obtain the temperature crack probability.

[0058] Step S4: According to the significant degree of the pixel value difference of the same suspected crack area in the road surface feature images before and after watering, the brightness crack probability and the temperature crack probability are adjusted to obtain the final crack probability of each suspected crack area; the final crack probability is used for road surface quality detection of the to-be-tested road.

[0059] The road surface crack has various types, and different types of cracks have different performances on the brightness information and the temperature information. The brightness information is sensitive to the surface texture, and the temperature information reflects the thermal characteristics inside the crack. The brightness difference and the temperature difference between the crack and the road surface before and after watering are larger, and the significant degree of the pixel value difference of the same suspected crack area in the road surface feature images before and after watering, i.e. the significant degree of the temperature change and the significant degree of the brightness change, is used to determine the importance of the road surface grayscale image and the road surface infrared image in the crack detection process, to dynamically adjust the proportion of the brightness crack probability and the temperature crack probability, to realize the dynamic fusion of the temperature difference information and the brightness information before and after watering, to obtain the final crack probability of the suspected crack area, and to improve the accuracy of the road surface crack detection.

[0060] Preferably, in some possible implementation manners of the embodiment of the present application, the method for obtaining the suspected crack area comprises: using the maximum inter-class variance method on the gray value of the pixel point in the road surface grayscale image before watering to obtain a segmentation threshold, and regarding the connected domain formed by the pixel points with the gray value less than the segmentation threshold as the suspected crack area of the road surface grayscale image before watering; regarding the road surface infrared image before and after watering and the road surface grayscale image after watering as analysis images; regarding the connected domain formed by the pixel points at the same coordinate position in each suspected crack area of the road surface grayscale image before watering in the analysis images as the suspected crack area of the analysis images. The maximum inter-class variance method is a known technology to those skilled in the art, and will not be introduced here.

[0061] It should be noted that, because the water seeps into the cracks after watering, the optical properties of the crack area are changed, and the infrared image is easily affected by external factors such as ambient temperature and wind speed, while the brightness information of the gray image directly reflects the change of the road surface texture, so the suspected crack area is determined by using the road surface gray image before watering. The crack is formed by micro-concave due to structural fracture, the incident light is absorbed or scattered in other directions, the light intensity reflected back to the camera is weakened, resulting in that the gray value of the crack in the road surface gray image is lower than that of the road surface, so the pixel points with gray value less than the segmentation threshold are selected to constitute the suspected crack area. At the same time, the road surface gray image before watering is one-to-one corresponding with the pixel points in the other three road surface feature images, so the suspected crack area in the road surface gray image before watering can be mapped to the other three road surface feature images, thereby obtaining the suspected crack area of the three road surface feature images.

[0062] It should be noted that, since the pixel points in each two road surface feature images are one-to-one corresponding, each suspected crack area in the road surface feature image refers to the suspected crack area with the same position in the road surface gray image before watering.

[0063] Preferably, in some possible implementation manners of the embodiment of the present application, the method for obtaining the brightness crack probability comprises: taking the mean value of the pixel values of all the pixel points in the suspected crack area as the pixel overall value; calculating the absolute value of the difference between the pixel overall values of the same suspected crack area in the road surface feature images before and after watering, to obtain the pixel difference of each suspected crack area in the road surface feature images; and performing normalization processing on the pixel difference of each suspected crack area in the road surface gray image, to obtain the brightness crack probability of each suspected crack area.

[0064] It should be noted that, as known, the crack has a larger brightness difference before and after watering than the road surface. The pixel overall value reflects the overall gray level of the suspected crack area in the road surface gray image or the overall temperature level of the suspected crack area in the road surface infrared image; the larger the pixel difference of the suspected crack area in the road surface feature image, the larger the brightness difference of the suspected crack area before and after watering, and the greater the possibility that the suspected crack area is identified as an actual crack through the brightness change characteristics before and after watering, and the greater the brightness crack probability. In the embodiment of the present application, the normalization processing is performed by using the Norm function, and other normalization methods such as function transformation and maximum-minimum normalization can also be selected, which are not limited herein.

[0065] Preferably, in some possible implementation manners of the embodiment of the present application, the method for obtaining the temperature crack probability comprises: the environmental data comprises temperature data and wind speed data; an environmental influence value is obtained according to the temperature data and the wind speed data; the pixel difference of each suspected crack area in the road surface infrared image is adjusted by using the environmental influence value, to obtain the actual temperature difference of each suspected crack area; and the actual temperature difference is normalized to obtain the temperature crack probability of each suspected crack area.

[0066] The greater the temperature and the wind speed, the faster the water evaporation rate, the more heat absorbed by the suspected crack area in unit time, the more significant the cooling effect, and the greater the environmental impact on water evaporation. Therefore, the temperature data and the wind speed data are positively correlated with the environmental impact value. In the embodiment of the present application, the product of the temperature data and the wind speed data is taken as the environmental impact value. The greater the environmental impact value, the more significant the cooling effect of the suspected crack area in unit time before and after watering, so that the actual temperature difference of the suspected crack area in the road surface infrared image is larger, thereby making the environmental impact value negatively correlated with the actual temperature difference. In the embodiment of the present application, the environmental impact value is negatively correlated and normalized, and the pixel difference of each suspected crack area in the road surface infrared image is weighted by using the processing result to obtain the actual temperature difference of each suspected crack area.

[0067] It is known that the temperature difference of a crack is greater than that of the road surface before and after watering. The greater the actual temperature difference, the greater the temperature difference of the suspected crack area before and after watering, and the greater the possibility of the suspected crack area being identified as an actual crack through the temperature change characteristics before and after watering, and the greater the temperature crack probability.

[0068] It should be noted that in the embodiment of the present application, the environmental impact value is taken as the index of the exponential function with a natural constant as the base number to achieve negative correlation and normalization processing. The normalization processing can also be performed by using the Norm function, or other normalization methods such as maximum and minimum normalization, which are not limited herein.

[0069] Preferably, in some possible implementation manners of the embodiment of the present application, the method for obtaining the final crack probability is as shown in Figure 2 which shows a flow chart of a method for obtaining a final crack probability provided by an embodiment of the present application. The method comprises the following steps:

[0070] Step S410: According to the significant degree of the pixel value difference of the same suspected crack area in the road surface feature image before and after watering, a feature weight coefficient is obtained.

[0071] Preferably, in some possible implementation manners of the embodiment of the present application, the method for obtaining the feature weight coefficient comprises: taking the ratio of the sum of the mean value, the standard deviation and a preset positive number of the pixel difference of all suspected crack areas in the road surface feature image as the feature significant degree; taking the corresponding feature significant degrees of the road surface grayscale image and the road surface infrared image as the grayscale feature significant degree and the infrared feature significant degree in turn; taking the infrared feature significant degree as the numerator and the sum of the infrared feature significant degree and the grayscale feature significant degree as the denominator to obtain the ratio as the feature weight coefficient.

[0072] In a specific implementation manner of the embodiment of the present application, the feature weight coefficient is expressed by formula as follows:

[0073]

[0074]

[0075]

[0076] wherein, is the infrared feature saliency; is the grayscale feature saliency; is the mean of pixel difference of all suspected crack regions in the infrared image of the road surface; and is the standard deviation of pixel difference of all suspected crack regions in the infrared image of the road surface; is the mean of pixel difference of all suspected crack regions in the grayscale image of the road surface; is the standard deviation of pixel difference of all suspected crack regions in the grayscale image of the road surface; is a preset positive number, which prevents the denominator from being zero and causing the fraction to be meaningless. In this embodiment, is set to 0.1, and the implementer can set it according to the specific circumstances.

[0077] It should be noted that the mean and the standard deviation of the pixel difference of all suspected crack regions in the road surface feature image reflect the average amplitude and the dispersion degree of the brightness change of all suspected crack regions before and after watering, respectively. If the feature saliency is greater, the temperature change or the brightness change of the suspected crack region before and after watering is more significant and consistent, and the authenticity of the crack is higher, which is suitable for identifying deep cracks; otherwise, the temperature change or the brightness change of the suspected crack region before and after watering is not significant, which is suitable for surface cracks. When is greater than , the temperature change feature of the crack is more significant, and the suspected crack region is more likely to be an interlayer separation crack, tends to 1, and whether the suspected crack region is an actual crack mainly depends on the temperature change before and after watering in the infrared image of the road surface; when is less than , the brightness change feature of the crack is more significant, and the suspected crack region is more likely to be a surface crack, tends to 0, and whether the suspected crack region is an actual crack mainly depends on the brightness change before and after watering in the grayscale image of the road surface.

[0078] Step S420: respectively taking the feature weight coefficient, the constant 1 and the difference between the feature weight coefficient as the weight of the temperature crack probability and the brightness crack probability of each suspected crack region, respectively, and performing weighted summation on the temperature crack probability and the brightness crack probability of each suspected crack region to obtain the final crack probability of each suspected crack region.

[0079] In a specific implementation of the embodiment of the present application, the final crack probability is expressed by the formula:

[0080]

[0081] In the formula, ZP is the final crack probability of each suspected crack region; is a feature weight coefficient; is the temperature crack probability of each suspected crack region; is the brightness crack probability of each suspected crack region. It should be noted that when the feature weight coefficient tends to 1, the determination of whether the suspected crack region is an actual crack mainly depends on the temperature change before and after watering, and the weight of the temperature crack probability should be greater, and the weight of the brightness crack probability should be smaller; when the feature weight coefficient tends to 0, the determination of whether the suspected crack region is an actual crack mainly depends on the brightness change before and after watering, and the weight of the temperature crack probability should be smaller, and the weight of the brightness crack probability should be greater. If the final crack probability is greater, the crack characteristics of the suspected crack region are more obvious, that is, the crack is more serious, and the possibility that the suspected crack region is an actual crack is greater.

[0082] In the embodiment of the present application, when the final crack probability of the suspected crack region is , the suspected crack region is a normal pavement region; when the final crack probability of the suspected crack region is , the suspected crack region is a small crack, and observation is recommended; when the final crack probability of the suspected crack region is , the suspected crack region is a potential expanding crack, and the staff needs to observe regularly to prevent the crack from expanding into a serious crack; when the final crack probability of the suspected crack region is , the suspected crack region is a serious crack, and the staff needs to repair immediately. Other embodiments can also divide the suspected crack region into cracks of different severity based on the final crack probability, which is not limited herein.

[0083] Embodiment 2

[0084] The present application provides a highway pavement quality detection system based on machine vision. Please refer to Figure 3 , which shows a system structure diagram of a highway pavement quality detection system based on machine vision provided by an embodiment of the present application. The system comprises:

[0085] The data acquisition module 510 is used to acquire pavement feature images of the to-be-detected highway before and after watering and environmental data of the to-be-detected highway position respectively. The pavement feature images include pavement grayscale images and pavement infrared images.

[0086] The brightness feature analysis module 520 is configured to acquire a suspected crack region in the pavement feature image; and acquire a brightness crack probability of each suspected crack region according to a gray value difference of the same suspected crack region in the pavement gray value images before and after watering.

[0087] The temperature feature analysis module 530 is configured to acquire a temperature crack probability of each suspected crack region according to a pixel value difference of the same suspected crack region in the pavement infrared images before and after watering and environmental data.

[0088] The pavement quality detection module 540 is configured to adjust the brightness crack probability and the temperature crack probability according to a significant degree of the pixel value difference of the same suspected crack region in the pavement feature images before and after watering, to acquire a final crack probability of each suspected crack region; and perform pavement quality detection on the to-be-detected road by using the final crack probability.

[0089] It should be noted that the device provided in the above embodiment is only exemplified by the division of the above functional modules, and in actual application, the above functions can be completed by different functional modules according to needs, that is, the internal structure of the computer device is divided into different functional modules to complete all or part of the functions described above. In addition, the highway pavement quality detection system and the highway pavement quality detection method provided in the above embodiment belong to the same concept, and the specific implementation process is described in the method embodiment, which will not be repeated here.

[0090] Embodiment 3

[0091] Figure 4 A computer device schematic diagram of a highway pavement quality detection device provided by an embodiment of the present application. As shown in the example, Figure 4 the computer device includes a memory 601, a processor 602, and a computer program 603 stored in the memory 601 and running on the processor 602, wherein when the processor 602 executes the computer program 603, the computer device can execute any of the above-described highway pavement quality detection methods based on machine vision.

[0092] In addition, the embodiment of the present application also protects a device, which can include a memory and a processor, wherein the memory stores executable program code, and the processor is configured to call and execute the executable program code to execute the highway pavement quality detection method based on machine vision provided by the embodiment of the present application.

[0093] The embodiment can divide the functions of the device into function modules according to the method examples described above. For example, each function module can be provided, or two or more functions can be integrated into one processing module. The integrated module can be implemented in the form of hardware. It should be noted that the division of the modules in the embodiment is illustrative, and is only a logical function division. In actual implementation, another division mode can be used.

[0094] It should be understood that the device provided by the embodiment is used to execute the machine vision-based highway pavement quality detection method described above, and thus the same effect as the implementation method described above can be achieved.

[0095] In the case of using an integrated unit, the device can include a processing module and a storage module. When the device is applied to equipment, the processing module can be used to control and manage the actions of the equipment. The storage module can be used to support the equipment to execute program codes and the like.

[0096] The processing module can be a processor or a controller, which can implement or execute various exemplary logical blocks, modules and circuits included in the disclosure of the present application. The processor can also be a combination of computing functions, such as one or more microprocessor combinations, a combination of digital signal processing (DSP) and microprocessor, and the like. The storage module can be a memory.

[0097] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description, and does not represent the advantages and disadvantages of the embodiments. The processes described in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multi-task processing and parallel processing are also possible or can be advantageous.

[0098] Each embodiment in the specification is described in a progressive manner, and the same or similar parts between each embodiment can be referred to each other. Each embodiment focuses on the differences from other embodiments.

[0099] The above only describes the preferred embodiments of the present application and does not limit the present application. Any modification, equivalent replacement, improvement and the like made within the principles of the present application shall be included in the protection scope of the present application.

Claims

1. A road surface quality detection method based on machine vision, characterized in that: The method includes: Obtaining road surface characteristic images of the road to be tested before and after watering, as well as environmental data of the location of the road to be tested; the road surface characteristic images include road surface grayscale images and road surface infrared images; Obtain suspected crack areas in the road feature image; obtain the brightness crack probability of each suspected crack area based on the grayscale difference of the same suspected crack area in the road grayscale image before and after watering; The temperature crack probability of each suspected crack area is obtained based on the difference in pixel values ​​of the same suspected crack area in the infrared images of the road surface before and after watering and the environmental data; Based on the significance of the difference in pixel values ​​in the same suspected crack area in the road surface feature images before and after watering, the brightness crack probability and the temperature crack probability are adjusted to obtain the final crack probability for each suspected crack area. The final crack probability is used to perform road surface quality inspection on the road under test. Obtain the brightness crack probability of each suspected crack area, including: The average pixel value of all pixels in the suspected crack area is taken as the overall pixel value; Calculate the absolute value of the difference between the overall pixel values ​​of the same suspected crack area in the road surface characteristic image before and after watering to obtain the pixel difference of each suspected crack area in the road surface characteristic image; Normalize the pixel difference of each suspected crack area in the road grayscale image to obtain the brightness crack probability of each suspected crack area; Obtain the temperature crack probability of each suspected crack area, including: Environmental data includes temperature data and wind speed data; environmental impact values ​​are obtained based on the temperature data and wind speed data; The environmental impact value is used to adjust the pixel difference of each suspected crack area in the pavement infrared image to obtain the actual temperature difference of each suspected crack area; the actual temperature difference is normalized to obtain the temperature crack probability of each suspected crack area.

2. The method for detecting road surface quality based on machine vision according to claim 1, characterized in that: Obtain the final crack probability for each suspected crack area, including: Obtain a feature weight coefficient based on the significance of the difference in pixel values ​​of the same suspected crack area in the road surface feature images before and after watering; The characteristic weight coefficient and the difference between the constant 1 and the characteristic weight coefficient are used as the weights of the temperature crack probability and the brightness crack probability of each suspected crack area respectively. The temperature crack probability and the brightness crack probability of each suspected crack area are weightedly summed to obtain the final crack probability of each suspected crack area.

3. The method for detecting road surface quality based on machine vision according to claim 2, characterized in that: Methods for obtaining feature weight coefficients include: The ratio of the mean, standard deviation and sum of the preset positive numbers of the pixel differences of all suspected crack areas in the road feature image is used as the feature significance. The feature significance of the road grayscale image and the road infrared image is recorded as the grayscale feature significance and the infrared feature significance respectively. The ratio of the infrared feature significance as the numerator and the sum of the infrared feature significance and the grayscale feature significance as the denominator is used as the feature weight coefficient.

4. The method for detecting road surface quality based on machine vision according to claim 1, characterized in that: Obtain suspected crack areas in the road feature image, including: The maximum inter-class variance method is used to obtain the segmentation threshold for the grayscale values ​​of the pixels in the grayscale image of the road surface before watering. The connected domain formed by the pixels with grayscale values ​​less than the segmentation threshold is regarded as the suspected crack area of ​​the grayscale image of the road surface before watering. The infrared images of the road surface before and after watering and the grayscale image of the road surface after watering are recorded as analysis images; the connected domain formed by the pixel points at the same coordinate position in the analysis image in each suspected crack area of ​​the road surface grayscale image before watering is used as the suspected crack area of ​​the analysis image.

5. The method for detecting road surface quality based on machine vision according to claim 1, characterized in that: Obtain the actual temperature difference of each suspected crack area, including: The environmental impact values ​​are negatively correlated and normalized, and the processing results are used to weight the pixel difference of each suspected crack area in the road surface infrared image to obtain the actual temperature difference of each suspected crack area.

6. The method for detecting road surface quality based on machine vision according to claim 1, characterized in that: Both temperature data and wind speed data are positively correlated with environmental impact values.

7. The method for detecting road surface quality based on machine vision according to claim 1, characterized in that: The pixels in every two road feature images correspond one to one.

8. A road surface quality detection system based on machine vision, characterized in that: The system includes: The data acquisition module is used to obtain the road surface characteristic images of the road to be tested before and after watering, as well as the environmental data of the location of the road to be tested; the road surface characteristic images include road surface grayscale images and road surface infrared images; The brightness feature analysis module is used to obtain suspected crack areas in the road surface feature image; based on the grayscale difference of the same suspected crack area in the road surface grayscale image before and after watering, the brightness crack probability of each suspected crack area is obtained; The temperature characteristic analysis module is used to obtain the temperature crack probability of each suspected crack area based on the difference in pixel values ​​of the same suspected crack area in the infrared images of the road surface before and after watering and the environmental data; The pavement quality detection module is used to adjust the brightness crack probability and temperature crack probability based on the significance of the difference in pixel values ​​in the same suspected crack area in the road feature images before and after watering, and obtain the final crack probability for each suspected crack area. The final crack probability is used to perform pavement quality detection on the road under test. Obtain the brightness crack probability of each suspected crack area, including: The average pixel value of all pixels in the suspected crack area is taken as the overall pixel value; Calculate the absolute value of the difference between the overall pixel values ​​of the same suspected crack area in the road surface characteristic image before and after watering to obtain the pixel difference of each suspected crack area in the road surface characteristic image; Normalize the pixel difference of each suspected crack area in the road grayscale image to obtain the brightness crack probability of each suspected crack area; Obtain the temperature crack probability of each suspected crack area, including: Environmental data includes temperature data and wind speed data; environmental impact values ​​are obtained based on the temperature data and wind speed data; The environmental impact value is used to adjust the pixel difference of each suspected crack area in the pavement infrared image to obtain the actual temperature difference of each suspected crack area; the actual temperature difference is normalized to obtain the temperature crack probability of each suspected crack area.

Images