Lightweight highway pavement aeolian sand change state recognition method and system

Abstract

The invention relates to the technical field of pavement aeolian sand states, in particular to a lightweight highway pavement aeolian sand change state recognition method and system. Firstly, an unmanned aerial vehicle is used for carrying a high-resolution camera to collect image data along a highway, and meanwhile meteorological data are collected and preprocessed. Converting the image into a grey-scale map, identifying an aeolian sand region through threshold judgment and morphological processing, and calculating a weighted gray value to judge the state of the aeolian sand region; then determining deformation and quantitative change characteristic factors of the aeolian sand, and analyzing correlation coefficients of the deformation and quantitative change characteristic factors and meteorological data; and constructing a deformation and quantitative change identification model, and outputting a deformation quantity and a quantitative quantity. And finally, determining the model weight according to the proportions of different types of aeolian sand, obtaining a change state index by combining the model output value and the weight, and judging the change state of the aeolian sand according to a preset threshold value, thereby realizing effective identification of the change state of the aeolian sand on the highway pavement. The method has the advantages that data acquisition is comprehensive, state recognition is accurate, influence factors can be mined, and a scientific basis is provided for road maintenance.

Description

Technical Field

[0001] The invention relates to the technical field of pavement wind-blown sand state, and in particular to a method and system for identifying the change state of wind-blown sand on a lightweight highway pavement. Background Art

[0002] In the field of highway construction and maintenance, highways located in sandy areas are often threatened by wind-blown sand. Wind-blown sand will accumulate on the road surface, which not only affects the normal use of the road and shortens its service life, but may also pose a serious hidden danger to driving safety. For example, the accumulation of wind-blown sand may reduce the friction coefficient of the road surface, making it easy for vehicles to slip when driving, especially on curves and steep slopes, increasing the risk of vehicle loss of control; the presence of a large amount of wind-blown sand will also cover road signs, interfere with the driver's vision, and cause traffic accidents.

[0003] Traditional methods of monitoring wind-blown sand have many limitations. Manual inspections are not only inefficient and difficult to achieve real-time and comprehensive monitoring of long-distance highways, but also limited by manpower and time, and cannot detect subtle changes in wind-blown sand in a timely manner. In addition, there is subjectivity in manual judgment of the state and degree of harm of wind-blown sand, and the judgment standards of different inspectors may be inconsistent, making it difficult to ensure the accuracy of monitoring results. With the development of science and technology, remote sensing technology has gradually been applied to highway monitoring. However, the resolution of early remote sensing images was low, making it difficult to accurately identify wind-blown sand areas on highway surfaces, and it was difficult to capture subtle features and changes in wind-blown sand. At the same time, meteorological factors have a great impact on the quality of remote sensing images. In windy and sandy weather or poor lighting conditions, the clarity and accuracy of the image will be seriously affected, resulting in a decrease in monitoring accuracy.

[0004] In terms of data analysis, previous studies on aeolian sand have focused on a single factor, such as focusing only on the amount of aeolian sand or analyzing its relationship with a certain meteorological factor, and have failed to comprehensively consider the impact of multiple factors on the changing state of aeolian sand. Moreover, there is a lack of effective models to accurately predict the changing trend of aeolian sand, making it difficult to take targeted protective measures in advance. Summary of the invention

[0005] In view of the shortcomings of the existing technology, the present invention proposes a method and system for identifying the changing state of wind-blown sand on lightweight highway pavement. Through image preprocessing, feature factor extraction, correlation analysis and model construction, a variety of factors are comprehensively considered to accurately identify the changing state of wind-blown sand, and provide accurate and timely information to highway maintenance departments so that effective protection and cleaning measures can be taken to ensure the safety, smoothness and service life of the highway.

[0006] The technical solution of the present invention is as follows:

[0007] One of the purposes of the present invention is to provide a method for identifying the changing state of wind-blown sand on a lightweight highway pavement, comprising:

[0008] Use drones equipped with high-resolution cameras to obtain image data along the highway and collect local meteorological data;

[0009] Calculate the pixel grayscale value of the image data, identify the wind-blown sand area, and obtain the weighted grayscale value of the wind-blown sand area on the road surface. By judging the weighted grayscale value, the wind-blown sand state is divided into static sand, mobile sand, and accumulated sand.

[0010] Analyze the deformation characteristic factors and quantitative characteristic factors of aeolian sand in different aeolian sand states, and use correlation coefficient analysis to calculate the correlation coefficient between the deformation characteristic factors and quantitative characteristic factors of aeolian sand and meteorological data;

[0011] According to the deformation characteristic factors of aeolian sand and the correlation between the deformation characteristic factors of aeolian sand and meteorological data, a deformation identification model of aeolian sand is established; according to the quantitative change characteristic factors of aeolian sand and the correlation between the quantitative change characteristic factors of aeolian sand and meteorological data, a quantitative change identification model of aeolian sand is established;

[0012] The weights of the aeolian sand deformation identification model and the aeolian sand quantitative change identification model are determined, and the final aeolian sand change state is determined by combining the output values ​​of the aeolian sand deformation identification model and the aeolian sand quantitative change identification model.

[0013] In a possible improved manner, the pixel grayscale value of the image data is obtained by converting the pre-processed color image data, and the pixel grayscale value calculation formula is:

[0014] I(x,y)=0.299R(x,y)+0.587G(x,y)+0.114B(x,y);

[0015] Among them, I(x,y) is the grayscale value at the coordinate (x,y), R(x,y) is the red sensitivity at the coordinate (x,y), G(x,y) is the green sensitivity at the coordinate (x,y), and B(x,y) is the blue sensitivity at the coordinate (x,y).

[0016] In a possible improved manner, the aeolian sand area is identified by gray value judgment and morphological processing; including:

[0017] Set the grayscale threshold range [I min ,I max ], where I min is the minimum threshold for determining the grayscale of aeolian sand, I max Maximum threshold for determining the grayscale of aeolian sand;

[0018] For the gray value I(x,y) at the coordinate (x,y), if I min ≤I(x,y)≤I max, it is preliminarily determined that the pixel belongs to the candidate area of ​​aeolian sand, where I(x,y) is the gray value at the coordinate (x,y);

[0019] Erosion and dilation operations are gradually performed on candidate aeolian sand areas to identify aeolian sand areas.

[0020] In a possible improved manner, the weighted gray value calculation formula of the aeolian sand area is:

[0021] I w (x,y)=I(x,y)×W(x,y);

[0022]

[0023] Among them, I w (x,y) is the weighted gray value of the pixel at the coordinate (x,y) in the aeolian sand area, I(x,y) is the gray value of the pixel at the coordinate (x,y) in the aeolian sand area, W(x,y) is the weight at the coordinate (x,y) in the aeolian sand area, (x0,y0) is the center coordinate of the highway surface, k is the weight adjustment parameter used to control the rate at which the weight changes with distance, and d(x,y) is the Euclidean distance from the coordinate (x,y) in the image to the center of the highway surface (x0,y0).

[0024] In a possible improved manner, the static sand is judged based on the standard deviation of the weighted gray value, that is, if the standard deviation is less than a set static sand standard deviation threshold, the aeolian sand area is judged to be static sand;

[0025] Shifting sand is determined based on the change rate of the weighted grayscale value of the aeolian sand area at adjacent time points. If the change rate is greater than the set shifting sand change rate threshold at multiple adjacent time points, the aeolian sand area is determined to be shifting sand.

[0026] Accumulated sand is judged based on the weighted gray value distribution and the rate of change of the weighted gray value in the aeolian sand area. The histogram of the weighted gray value in the aeolian sand area is obtained by calculation. If the histogram shows an obvious single peak and the peak gray value gradually increases, and the rate of change is less than the threshold value of the rate of change of mobile sand, the aeolian sand area is judged to be accumulated sand.

[0027] In a possible improved manner, the deformation characteristic factor includes a surface profile change factor and a slope change factor;

[0028] The surface profile variation factor calculation formula is:

[0029]

[0030] Among them, R edgeis the surface profile variation factor, P1 and P2 are the edge profile perimeters of the surface profile obtained at time t1 and t2, respectively;

[0031] The slope change factor calculation formula is:

[0032]

[0033] Among them, C flat is the surface flatness variation factor; σ1 and σ2 are the weighted gray value standard deviations of the aeolian sand area;

[0034] The quantitative change characteristic factors include area change factor and gray value sum change factor;

[0035] The area change factor calculation formula is:

[0036] ΔA=A2-A1;

[0037] Among them, ΔA is the area change factor, A1 and A2 are the areas of aeolian sand obtained at time t1 and t2 respectively;

[0038] The calculation formula of the total gray value change factor is:

[0039] ΔI S =I S2 -I S1 ;

[0040] Among them, ΔI S is the total gray value change factor, I S1 and I S2 are the weighted gray value sums of all pixels in the aeolian sand area at different times.

[0041] In a possible improved manner, the calculation formula for the correlation coefficient between the deformation characteristic factor and the quantitative characteristic factor of the aeolian sand and the meteorological data is:

[0042]

[0043] Among them, r ab is the correlation coefficient between a and b, a is a certain meteorological data, b is one of the deformation characteristic factors or one of the quantitative characteristic factors of aeolian sand, a i and b i are the values ​​of a and b in the i-th sample, and are the means of a and b respectively, and n is the number of samples.

[0044] In a possible improved manner, the aeolian sand deformation identification model is:

[0045]

[0046] Where D is the aeolian sand shape variable, r i is the correlation coefficient between meteorological factors and deformation characteristic factors, F 形变,i is the corresponding deformation characteristic factor, m is the number of meteorological factors involved in modeling;

[0047] The aeolian sand quantitative change identification model is:

[0048]

[0049] Among them, Q is the variable of wind-blown sand, r j is the correlation coefficient between meteorological factors and quantitative characteristic factors, F 量变,j is the corresponding quantitative characteristic factor, and n is the number of meteorological factors involved in modeling.

[0050] In a possible improved manner, the final method for determining the change state of aeolian sand includes:

[0051] Obtain the change status index of aeolian sand, namely:

[0052] S=ω D ×D+ω Q ×Q;

[0053] Among them, S is the index of wind-blown sand change state, ω D and ω Q are the weights of the aeolian sand deformation identification model and the weights of the aeolian sand quantitative change identification model, respectively. D and Q are the deformation and quantitative variables of aeolian sand, respectively.

[0054] Preset the threshold range of wind-blown sand change state [S min ,S max ];

[0055] Determine the state of aeolian sand based on the threshold range;

[0056] Among them, when S≤S min When , it is determined that the aeolian sand is in a stable state;

[0057] When S min <S≤S max , aeolian sand is in a slowly changing state;

[0058] When S>S max , aeolian sand is in a state of rapid change, and the state type of aeolian sand should be re-evaluated.

[0059] Another object of the present invention is to provide a system for identifying the changing state of wind-blown sand on a lightweight highway pavement, comprising:

[0060] The data acquisition module is used to use a drone equipped with a high-resolution camera to acquire image data along the highway and collect local meteorological data;

[0061] The module for wind-blown sand area identification, weighted gray value calculation and state classification is used to calculate the pixel gray value of the image data, identify the wind-blown sand area, and obtain the weighted gray value of the wind-blown sand area on the road surface. By judging the weighted gray value, the wind-blown sand state is divided into static sand, mobile sand, and accumulated sand.

[0062] The characteristic factor analysis and correlation coefficient calculation module is used to analyze the deformation characteristic factors and quantitative characteristic factors of aeolian sand in different aeolian sand states, and calculate the correlation coefficient between the deformation characteristic factors and quantitative characteristic factors of aeolian sand and meteorological data by using correlation coefficient analysis;

[0063] The identification model establishment module is used to establish a deformation identification model of aeolian sand according to the deformation characteristic factors of aeolian sand and the correlation between the deformation characteristic factors of aeolian sand and meteorological data; and to establish a quantitative change identification model of aeolian sand according to the quantitative change characteristic factors of aeolian sand and the correlation between the quantitative change characteristic factors of aeolian sand and meteorological data;

[0064] The wind-blown sand change state determination module is used to determine the weights of the wind-blown sand deformation recognition model and the wind-blown sand quantity change recognition model, and determine the final wind-blown sand change state in combination with the output values ​​of the wind-blown sand deformation recognition model and the wind-blown sand quantity change recognition model.

[0065] The beneficial effects brought by the technical solution provided by the embodiment of the present application include at least the following:

[0066] Beneficial effects:

[0067] 1. The high-resolution image data obtained by UAVs and the meteorological data collected by meteorological stations or meteorological sensors comprehensively consider the various factors affecting the change of aeolian sand, making the identification results more comprehensive and accurate. At the same time, the deformation characteristic factors and quantitative characteristic factors of aeolian sand are analyzed, and combined with the correlation of meteorological data, the changing state of aeolian sand can be more accurately reflected.

[0068] 2. The state of aeolian sand is divided into static sand, mobile sand and accumulated sand. By analyzing the statistical characteristics and change rules of weighted gray values, the specific state of aeolian sand can be accurately determined, providing a scientific basis for the maintenance and management of highway pavement. By analyzing image data at different times, the changes of aeolian sand can be monitored in real time, and the dynamic changes of aeolian sand can be discovered in time, which helps to take corresponding measures in advance to reduce the impact of aeolian sand on highway pavement.

[0069] 3. The constructed aeolian sand deformation and quantitative change identification model combines the deformation and quantitative change characteristic factors with meteorological data based on the correlation coefficient, and outputs the deformation variable and quantitative variable respectively, accurately reflecting the degree of change of aeolian sand. The model weight is determined by counting the proportion of different types of aeolian sand, and the weighted sum is used to obtain the aeolian sand change state index. The stability, slow change or rapid change state of aeolian sand is judged based on the threshold, and the state type can be re-evaluated and updated when it changes rapidly, so as to realize accurate monitoring and prediction of the change state of aeolian sand.

[0070] 4. The present invention can timely and accurately grasp the changing state of wind-blown sand, and provide a basis for decision-making for highway maintenance departments. When the wind-blown sand is in a stable state, regular inspections are reasonably arranged; when it is in a slowly changing state, the monitoring frequency is increased; when it is in a rapidly changing state, cleaning and protection measures are taken in a timely manner to ensure the stability of the highway pavement structure and driving safety, extend the service life of the highway, and reduce maintenance costs. It has important practical application significance. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] Figure 1 This is a schematic diagram of the overall process of the method for identifying the changing state of wind-blown sand on lightweight highway pavement;

[0072] Figure 2 A flow chart of the steps of a method S100 for identifying the changing state of wind-blown sand on a lightweight highway pavement;

[0073] Figure 3 The flowchart of the step S200 of the method for identifying the change state of wind-blown sand on a lightweight highway pavement;

[0074] Figure 4 The step flow chart of the method S300 for identifying the change state of wind-blown sand on a lightweight highway pavement;

[0075] Figure 5 A flowchart of the step S400 of the method for identifying the change state of wind-blown sand on a lightweight highway pavement;

[0076] Figure 6 The flowchart of step S500 is a method for identifying the changing state of wind-blown sand on a lightweight highway pavement. DETAILED DESCRIPTION

[0077] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0078] The impact of aeolian sand on the road surface is a key issue in the construction and maintenance of roads in deserts or areas with heavy wind and sand. The continuous accumulation and movement of aeolian sand will change the structure of the road surface, reduce the bearing capacity of the road surface, and increase the risk of driving safety. Traditional aeolian sand monitoring methods often rely on manual field surveys, which are inefficient and inaccurate, and it is difficult to achieve real-time and dynamic monitoring of the state of aeolian sand along large areas of roads. In addition, although some existing monitoring technologies can obtain some aeolian sand information, they cannot fully and accurately identify the changing state of aeolian sand, and cannot provide reliable data support for highway maintenance and management.

[0079] To solve the above problems, please refer to Figure 1 , which shows a method and system for identifying the change state of wind-blown sand on a lightweight highway pavement provided by an embodiment of the present invention, the method comprising:

[0080] S100: Use drones equipped with high-resolution cameras to obtain image data along the highway and collect local meteorological data.

[0081] S200: Calculate the pixel grayscale value of the image data, identify the wind-blown sand area, and obtain the weighted grayscale value of the wind-blown sand area on the road surface. By judging the weighted grayscale value, the wind-blown sand state is divided into static sand, flowing sand, and accumulated sand.

[0082] S300: Analyze the deformation characteristic factors and quantitative change characteristic factors of aeolian sand in different aeolian sand states, and use correlation coefficient analysis to calculate the correlation coefficient between the deformation characteristic factors and quantitative change characteristic factors of aeolian sand and meteorological data.

[0083] S400: establishing a deformation identification model of aeolian sand according to deformation characteristic factors of aeolian sand and the correlation between the deformation characteristic factors of aeolian sand and meteorological data; establishing a quantitative change identification model of aeolian sand according to quantitative change characteristic factors of aeolian sand and the correlation between the quantitative change characteristic factors of aeolian sand and meteorological data.

[0084] S500: Determine the weights of the aeolian sand deformation recognition model and the aeolian sand quantitative change recognition model, and determine the final aeolian sand change state by combining the output values ​​of the aeolian sand deformation recognition model and the aeolian sand quantitative change recognition model.

[0085] S100 uses drones equipped with high-resolution cameras to collect image data along the highway and collect local meteorological data, taking multiple factors into consideration during the collection process to ensure data quality.

[0086] Please refer to Figure 2 , which shows a flowchart of an exemplary method S100 for identifying the change state of wind-blown sand on a lightweight highway pavement of the present application, and its contents include:

[0087] S110: UAV image data acquisition and meteorological data collection.

[0088] Plan the flight path of the drone based on the length of the highway, the complexity of the terrain, and the endurance of the equipment to ensure that all target areas are covered. Use the drone equipped with a high-resolution camera to shoot along the highway according to the pre-set flight route and altitude. During the shooting process, consider factors such as lighting conditions and weather conditions to ensure that clear and accurate image data is obtained.

[0089] In a possible implementation, visible light and near-infrared band images are collected to enhance the contrast between aeolian sand and the road surface.

[0090] At the same time, record the time, location and other information of the shooting for subsequent data processing and analysis.

[0091] Meteorological data is collected through weather stations or weather sensors to collect local meteorological data, including wind speed, wind direction, temperature, humidity, precipitation, etc. Ensure that the timestamp and geographic location of the meteorological data are aligned with the image data.

[0092] S120: Preprocessing of image data and meteorological data.

[0093] Image data preprocessing includes:

[0094] The acquired image data is subjected to radiation correction to eliminate the influence of solar radiation, atmospheric conditions and other factors on the image brightness, so that the grayscale value of the image can more accurately reflect the true reflection characteristics of the ground object.

[0095] Exemplarily, the dark object subtraction method is used to perform radiation correction, by calculating the average gray value of dark object pixels in the image and subtracting it from the gray values ​​of all pixels, thereby obtaining corrected image data.

[0096] Perform geometric correction to correct the geometric deformation of the image caused by factors such as the flight attitude of the drone and terrain undulations, so that the image matches the actual geographic coordinate system.

[0097] For example, the control points are used to perform polynomial fitting and geometric correction of the image to ensure that linear features such as roads remain straight in the image.

[0098] Meteorological data preprocessing includes:

[0099] The collected meteorological data are cleaned to remove outliers and erroneous data.

[0100] For example, if the temperature value recorded by a certain meteorological sensor at a certain moment is obviously beyond the normal range, it may be caused by sensor failure or other reasons, and it needs to be eliminated.

[0101] Time series interpolation is performed on meteorological data to fill in missing data points and ensure the integrity and continuity of meteorological data.

[0102] Exemplarily, a linear interpolation method is used to estimate the meteorological data value at the missing time based on the meteorological data at adjacent time points.

[0103] S200 first converts the pre-processed color image into a grayscale image, identifies the aeolian sand area by setting the grayscale threshold and morphological processing, calculates its weighted grayscale value, and then determines the state of the aeolian sand based on the statistical characteristics and change rules of the weighted grayscale value.

[0104] Please refer to Figure 2 , which shows a flowchart of an exemplary lightweight highway pavement wind-blown sand change state identification method S200 of the present application, and its contents include:

[0105] S210: Convert the pre-processed color image data into a grayscale image.

[0106] The preprocessed color image data is converted into a grayscale image for subsequent pixel grayscale value calculation.

[0107] In a possible implementation, a weighted average method is used to assign different weights to the three color channels of red R, green G, and blue B according to the difference in sensitivity of the human eye to different color channels. The calculation formula is:

[0108] I(x,y)=0.299R(x,y)+0.587G(x,y)+0.114B(x,y);

[0109] Among them, I(x,y) is the grayscale value at the coordinate (x,y), R(x,y) is the red sensitivity at the coordinate (x,y), G(x,y) is the green sensitivity at the coordinate (x,y), and B(x,y) is the blue sensitivity at the coordinate (x,y).

[0110] Through this formula, each pixel in the color image is converted into a corresponding grayscale value to generate a grayscale image. The processed image retains the brightness information about the distribution of aeolian sand in the original image, while simplifying the data processing process and improving the subsequent calculation efficiency.

[0111] S220: Identify the aeolian sand area through grayscale value judgment and morphological processing, and calculate the pixel grayscale value of the aeolian sand area through weighted average method.

[0112] In order to preliminarily identify the aeolian sand area, a grayscale threshold range [I min ,I max], the threshold range is determined based on the statistical analysis of a large amount of aeolian sand image data and actual engineering experience. Traversing the entire grayscale image, for the grayscale value I(x,y) at the coordinate (x,y), if I min ≤I(x,y)≤I max , it is preliminarily determined that the pixel belongs to the candidate area of ​​aeolian sand.

[0113] Morphological processing is performed on the candidate regions of aeolian sand, including erosion and dilation operations, to accurately identify the aeolian sand regions. The erosion operation uses structural elements (such as squares, circles, etc.) to erode the initially identified aeolian sand regions. The purpose is to remove isolated noise points in the aeolian sand regions and tiny protrusions on the edges, making the boundaries of the aeolian sand regions clearer and more regular. The dilation operation is to dilate the processed regions after the erosion operation. The dilation operation can fill the holes caused by erosion in the aeolian sand regions, connect some originally disconnected regions, and more accurately restore the actual shape of the aeolian sand regions.

[0114] After identifying the aeolian sand area, in order to more accurately reflect the characteristics of the aeolian sand area, the weighted average method is used to calculate the pixel grayscale value of the aeolian sand area.

[0115] Considering that pixels at different positions in the aeolian sand area contribute differently to the overall features, a weight is assigned to each pixel.

[0116] Assume that the gray value of the pixel at the coordinate (x, y) in the aeolian sand area is I(x, y), and its weight is W(x, y). Then the weighted gray value of the pixel is I w The calculation formula for (x,y) is:

[0117] I w (x,y)=I(x,y)×W(x,y).

[0118] The weight is determined based on the distance between the pixel and the center of the road surface. Pixels closer to the center of the road surface are given higher weights because they have a more direct impact on the wind-blown sand on the road; pixels farther away have relatively lower weights. Therefore, the weight W(x,y) of the pixel at the coordinate (x,y) in the wind-blown sand area is calculated as:

[0119]

[0120] Among them, (x0, y0) is the coordinate of the center of the highway surface, k is the weight adjustment parameter used to control the rate at which the weight changes with distance, and d(x, y) is the Euclidean distance from the coordinate (x, y) in the image to the center of the highway surface (x0, y0).

[0121] By traversing the pixels of the entire aeolian sand area, calculating the weighted gray value of each pixel, and performing statistical analysis on all weighted gray values, the weighted gray value of the aeolian sand area is obtained.

[0122] S230: For the determined aeolian sand area, the state of the aeolian sand is determined by analyzing the statistical characteristics and change patterns of the weighted gray values.

[0123] Static sand judgment:

[0124] The weighted grayscale value of the static sand area is relatively stable and changes little within a certain period of time. Calculate the weighted grayscale value mean of the aeolian sand area at multiple time points and standard deviation σ s .

[0125] In the data collected at n time points, the weighted gray value of the area at the i-th time point is I wi , then the mean calculation formula is:

[0126]

[0127] The formula for calculating standard deviation is:

[0128]

[0129] If the standard deviation σ s If it is smaller than the set static sand standard deviation threshold, the aeolian sand area can be judged as static sand.

[0130] For example, after a large amount of data statistics, it is found that the standard deviation of the weighted grayscale value of the static sand area generally does not exceed 10, so the static sand standard deviation threshold can be set to 10.

[0131] Quicksand judgment:

[0132] The weighted grayscale value of the moving sand area will show a large dynamic change due to the movement of sand particles. The weighted grayscale value of the wind-blown sand area at adjacent time points is calculated to determine whether it is moving sand. Suppose the weighted grayscale values ​​of adjacent time points t1 and t2 are I wt1 and I wt2 , then the calculation formula of the rate of change r is:

[0133]

[0134] If the change rate r is greater than the set quicksand change rate threshold at multiple adjacent time points, the area is judged to be quicksand.

[0135] For example, after observing the quicksand area, it is found that the weighted gray value change rate is usually greater than 0.2, so the quicksand change rate threshold can be set to 0.2.

[0136] Accumulated sand judgment:

[0137] The weighted grayscale value characteristics of the accumulated sand area are between static sand and mobile sand. Its grayscale value may change with the increase of the accumulation amount, but the change speed is slower than that of mobile sand. It can be judged by analyzing the distribution of the weighted grayscale values ​​of the aeolian sand area. Calculate the histogram of the weighted grayscale values ​​of the aeolian sand area and observe its distribution pattern. If the histogram shows an obvious single peak and the peak grayscale value gradually increases, it means that the increase in the accumulation amount causes the grayscale value to change, and the change rate is less than the change rate threshold of the mobile sand, then the area can be judged as accumulated sand.

[0138] S300 determines the deformation and quantitative characteristic factors of aeolian sand, and uses the correlation coefficient analysis algorithm to calculate the correlation coefficient between these characteristic factors and meteorological data.

[0139] Please refer to Figure 4 , which shows a flowchart of an exemplary method S300 for identifying the change state of wind-blown sand on a lightweight highway pavement of the present application, and its contents include:

[0140] S310: Determine deformation characteristic factors.

[0141] The deformation characteristic factors include surface profile change factor and slope change factor.

[0142] The surface profile change factor is to extract the edge profile of the aeolian sand area through the edge detection algorithm. The difference between the edge profiles of the aeolian sand during two consecutive monitorings is calculated to obtain the surface profile change characteristic factor. For each aeolian sand state, the edge profile perimeters of the surface profile obtained at time t1 and t2 are P1 and P2 respectively, and the surface profile change factor R edge The calculation formula is:

[0143]

[0144] The surface smoothness variation factor is used to approximate the smoothness of the aeolian sand surface by the difference in the weighted grayscale values ​​of the aeolian sand area. The aeolian sand area is divided into multiple small sub-areas. For each sub-area, the standard deviation σ of its weighted grayscale value is calculated. Suppose the standard deviations of the weighted grayscale values ​​of a sub-area are σ1 and σ2 respectively, and the surface smoothness variation factor C is flat for:

[0145]

[0146] S320: Determine the quantitative change characteristic factor.

[0147] The quantitative change characteristic factors include the area change factor and the gray value sum change factor.

[0148] The area change factor uses image segmentation technology to obtain the number of pixels in the aeolian sand area at different times. According to the resolution of the image, the number of pixels is converted into the actual area. Assuming that the area of ​​the aeolian sand area obtained at time t1 and t2 is A1 and A2 respectively, the area change factor ΔA is:

[0149] ΔA=A2-A1.

[0150] The total gray value variation factor is the weighted gray value sum of all pixels in the aeolian sand area at different times. S1 and I S2 Subtract to obtain. Gray value total change factor ΔI S The calculation formula is:

[0151] ΔI S =I S2 -I S1 .

[0152] S330: Calculate the correlation coefficient between the deformation characteristic factor, quantitative characteristic factor and meteorological data of aeolian sand using the correlation coefficient analysis algorithm.

[0153] The correlation coefficient calculation formula is:

[0154]

[0155] Among them, r ab is the correlation coefficient between a and b, a is a certain meteorological data, b is one of the deformation characteristic factors or one of the quantitative characteristic factors of aeolian sand, a i and b i are the values ​​of a and b in the i-th sample, and are the means of a and b respectively, and n is the number of samples.

[0156] According to this formula, the correlation coefficients between meteorological data and various deformation characteristic factors and quantitative change characteristic factors are calculated. By analyzing the correlation coefficients, the meteorological factors that have a significant impact on the deformation and quantitative change of aeolian sand are found, providing a key basis for the subsequent establishment of a prediction model for the change state of aeolian sand.

[0157] Based on the deformation and quantitative change characteristic factors and their correlation coefficients with meteorological data, S400 constructs a deformation identification model for aeolian sand and a quantitative change identification model for aeolian sand, and outputs the deformation and quantitative variables of aeolian sand.

[0158] Please refer to Figure 5 , which shows a flowchart of an exemplary lightweight highway pavement wind-blown sand change state identification method S400 of the present application, and its contents include:

[0159] S410: Establish a wind-blown sand deformation identification model.

[0160] The deformation variable D is composed of the deformation characteristic factor, that is, the surface profile change factor R edge and surface flatness variation factor C flat The weighted combination of relevant meteorological factors makes the aeolian sand deformation identification model:

[0161]

[0162] Where D is the aeolian sand shape variable, r i is the correlation coefficient between meteorological factors and deformation characteristic factors, F 形变,i is the corresponding deformation characteristic factor, and m is the number of meteorological factors involved in modeling.

[0163] S420: Establish a model for identifying the quantitative changes of aeolian sand.

[0164] The quantitative variable Q is composed of the quantitative characteristic factors, namely the area change factor ΔA and the gray value total change factor ΔI S The weighted combination of relevant meteorological factors makes the identification model of aeolian sand quantitative change:

[0165]

[0166] Among them, Q is the variable of wind-blown sand, r j is the correlation coefficient between meteorological factors and quantitative characteristic factors, F 量变,j is the corresponding quantitative characteristic factor, and n is the number of meteorological factors involved in modeling.

[0167] In the aeolian sand deformation identification model and the aeolian sand quantitative change identification model, the deformation characteristic factor and the correlation coefficient between the deformation characteristic factor and the meteorological data, the quantitative change characteristic factor and the correlation coefficient between the quantitative change characteristic factor and the meteorological data are input respectively, and the deformation variable and the quantitative variable are output respectively. The deformation variable reflects the comprehensive index of the degree of aeolian sand deformation, and the quantitative variable reflects the comprehensive index of the degree of aeolian sand accumulation or erosion.

[0168] S500: The proportions of different types of aeolian sand are counted to determine the weights of the two models, corresponding characteristic factors and correlation coefficients are input to obtain model output values, aeolian sand change state index is obtained by weighted summation, and the change state of aeolian sand is determined by comparing with a preset threshold.

[0169] Please refer to Figure 6 , which shows a flowchart of an exemplary lightweight highway pavement wind-blown sand change state identification method S500 of the present application, and its contents include:

[0170] S510: Counting the proportions of different types of aeolian sand, and determining the weights of aeolian sand deformation identification model and aeolian sand quantitative change identification model.

[0171] The entire road surface monitoring area is traversed, and according to the classification results of the aeolian sand state in step S200, that is, static sand, mobile sand, and accumulated sand, the area and mass proportion of each type of aeolian sand in the monitoring area are counted. According to the proportion of different types of aeolian sand and the importance of the impact of each type of aeolian sand on the road surface, the weights of the aeolian sand deformation identification model and the aeolian sand quantitative change identification model are determined.

[0172] In a possible implementation, mobile sand and accumulated sand have a relatively greater impact on the road pavement structure and driving safety, and are assigned a relatively higher weight; static sand has a relatively smaller impact, and is accordingly assigned a lower weight.

[0173] S520: Obtain output values ​​of the aeolian sand deformation identification model and the aeolian sand quantitative change identification model.

[0174] The deformation characteristic factor currently monitored, that is, the surface profile change factor R edge and surface flatness variation factor C flat The correlation coefficient between it and meteorological data is input into the aeolian sand deformation identification model to obtain the aeolian sand deformation variable D.

[0175] The quantitative characteristic factors currently monitored, namely the area change factor ΔA and the gray value sum change factor ΔI S The correlation coefficient between the wind-blown sand and the meteorological data is input into the quantitative change identification model of the wind-blown sand to obtain the quantitative variable Q of the wind-blown sand.

[0176] S530: Determine the final change state of aeolian sand.

[0177] Combining the output values ​​of the aeolian sand deformation identification model and the aeolian sand quantity change identification model and the corresponding weights, the final aeolian sand change state index is determined by weighted summation. The calculation formula for the aeolian sand change state index is:

[0178] S=ω D ×D+ω Q ×Q;

[0179] Among them, S is the index of wind-blown sand change state, ω D and ω Q are the weights of the aeolian sand deformation identification model and the weights of the aeolian sand quantitative change identification model, D and Q are the deformation variable and quantitative variable of aeolian sand respectively.

[0180] Preset the threshold range of wind-blown sand change state [S min ,S max ], and judge the change state of aeolian sand according to the value of change state index S:

[0181] When S≤S minWhen the wind-blown sand is in a stable state, it is judged that the wind-blown sand has little impact on the road surface, that is, the static sand, flowing sand or accumulated sand state judged before is maintained, and the road surface condition is relatively stable.

[0182] When S min <S≤S max , aeolian sand is in a slowly changing state, recording the current aeolian sand state type and its slowly changing trend.

[0183] When S>S max , the aeolian sand is in a rapidly changing state, and the aeolian sand state type is re-evaluated. If the current state is static sand, and the change trends of D and Q meet the characteristics of mobile sand or accumulated sand, the aeolian sand state type is updated; similarly, similar judgments and updates are made for mobile sand and accumulated sand.

[0184] The following is a system embodiment of the present application. For details not described in detail in the system embodiment, reference can be made to the corresponding records in the above method embodiment, and they will not be repeated herein.

[0185] Lightweight highway pavement wind-blown sand change state identification system, including:

[0186] The data acquisition module is used to use a drone equipped with a high-resolution camera to acquire image data along the highway and collect local meteorological data;

[0187] The module for wind-blown sand area identification, weighted gray value calculation and state classification is used to calculate the pixel gray value of the image data, identify the wind-blown sand area, and obtain the weighted gray value of the wind-blown sand area on the road surface. By judging the weighted gray value, the wind-blown sand state is divided into static sand, mobile sand, and accumulated sand.

[0188] The characteristic factor analysis and correlation coefficient calculation module is used to analyze the deformation characteristic factors and quantitative characteristic factors of aeolian sand in different aeolian sand states, and calculate the correlation coefficient between the deformation characteristic factors and quantitative characteristic factors of aeolian sand and meteorological data by using correlation coefficient analysis;

[0189] The identification model establishment module is used to establish a deformation identification model of aeolian sand according to the deformation characteristic factors of aeolian sand and the correlation between the deformation characteristic factors of aeolian sand and meteorological data; and to establish a quantitative change identification model of aeolian sand according to the quantitative change characteristic factors of aeolian sand and the correlation between the quantitative change characteristic factors of aeolian sand and meteorological data;

[0190] The wind-blown sand change state determination module is used to determine the weights of the wind-blown sand deformation recognition model and the wind-blown sand quantity change recognition model, and determine the final wind-blown sand change state in combination with the output values ​​of the wind-blown sand deformation recognition model and the wind-blown sand quantity change recognition model.

[0191] The basic principles of the present application are described above in conjunction with specific embodiments. However, it should be noted that the advantages, strengths, effects, etc. mentioned in the present application are only examples and not limitations, and it cannot be considered that these advantages, strengths, effects, etc. are required by each embodiment of the present application. In addition, the specific details disclosed above are only for the purpose of illustration and ease of understanding, not for limitation, and the above details do not limit the present application to being implemented by adopting the above specific details.

[0192] The block diagrams of the devices, apparatuses, equipment, and systems involved in this application are only illustrative examples and are not intended to require or imply that they must be connected, arranged, and configured in the manner shown in the block diagram. As will be appreciated by those skilled in the art, these devices, apparatuses, equipment, and systems can be connected, arranged, and configured in any manner. Words such as "including", "comprising", "having", etc. are open words, referring to "including but not limited to", and can be used interchangeably with them. The words "or" and "and" used here refer to the words "and / or" and can be used interchangeably with them, unless the context clearly indicates otherwise. The words "such as" used here refer to the phrase "such as but not limited to", and can be used interchangeably with them.

[0193] It should also be noted that in the apparatus, device and method of the present application, each component or each step can be decomposed and / or recombined. Such decomposition and / or recombination should be regarded as equivalent solutions of the present application.

[0194] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other aspects without departing from the scope of the present application. Therefore, the present application is not intended to be limited to the aspects shown herein, but rather to the widest scope consistent with the principles and novel features of the present invention.

[0195] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A method for identifying the changing state of wind-blown sand on a lightweight highway pavement, characterized in that: include: Use drones equipped with high-resolution cameras to obtain image data along the highway and collect local meteorological data; Calculate the pixel grayscale value of the image data, identify the wind-blown sand area, and obtain the weighted grayscale value of the wind-blown sand area on the road surface. By judging the weighted grayscale value, the wind-blown sand state is divided into static sand, mobile sand, and accumulated sand. Analyze the deformation characteristic factors and quantitative change characteristic factors of aeolian sand in different aeolian sand states, and use correlation coefficient analysis to calculate the correlation coefficient between the deformation characteristic factors and quantitative change characteristic factors of aeolian sand and meteorological data; According to the deformation characteristic factors of aeolian sand and the correlation between the deformation characteristic factors of aeolian sand and meteorological data, a deformation identification model of aeolian sand is established; according to the quantitative change characteristic factors of aeolian sand and the correlation between the quantitative change characteristic factors of aeolian sand and meteorological data, a quantitative change identification model of aeolian sand is established; The weights of the aeolian sand deformation identification model and the aeolian sand quantitative change identification model are determined, and the final aeolian sand change state is determined by combining the output values ​​of the aeolian sand deformation identification model and the aeolian sand quantitative change identification model.

2. The method for identifying the change state of wind-blown sand on lightweight highway pavement according to claim 1 is characterized in that: The pixel grayscale value of the image data is obtained by converting the pre-processed color image data. The pixel grayscale value calculation formula is: I(x,y)=0.299R(x,y)+0.587G(x,y)+0.114B(x,y); Among them, I(x,y) is the grayscale value at the coordinate (x,y), R(x,y) is the red sensitivity at the coordinate (x,y), G(x,y) is the green sensitivity at the coordinate (x,y), and B(x,y) is the blue sensitivity at the coordinate (x,y).

3. The method for identifying the change state of wind-blown sand on lightweight highway pavement according to claim 2 is characterized in that: The aeolian sand area is identified by gray value judgment and morphological processing; including: Set the grayscale threshold range [I min ,I max ], where I min is the minimum threshold for determining the grayscale of aeolian sand, I max Maximum threshold for determining the grayscale of aeolian sand; For the gray value I(x,y) at the coordinate (x,y), if I min ≤I(x,y)≤I max , it is preliminarily determined that the pixel belongs to the candidate area of ​​aeolian sand, where I(x,y) is the gray value at the coordinate (x,y); Erosion and dilation operations are gradually performed on candidate aeolian sand areas to identify aeolian sand areas.

4. The method for identifying the change state of wind-blown sand on lightweight highway pavement according to claim 1 is characterized in that: The weighted gray value calculation formula of the aeolian sand area is: I w (x,y)=I(x,y)×W(x,y); Among them, I w (x,y) is the weighted gray value of the pixel at the coordinate (x,y) in the aeolian sand area, I(x,y) is the gray value of the pixel at the coordinate (x,y) in the aeolian sand area, W(x,y) is the weight at the coordinate (x,y) in the aeolian sand area, (x0,y0) is the center coordinate of the highway surface, k is the weight adjustment parameter used to control the rate at which the weight changes with distance, and d(x,y) is the Euclidean distance from the coordinate (x,y) in the image to the center of the highway surface (x0,y0).

5. The method for identifying the change state of wind-blown sand on lightweight highway pavement according to claim 4 is characterized in that: The static sand is determined based on the standard deviation of the weighted gray value. If the standard deviation is less than the set static sand standard deviation threshold, the aeolian sand area is determined to be static sand. Shifting sand is determined based on the change rate of the weighted grayscale value of the aeolian sand area at adjacent time points. If the change rate is greater than the set shifting sand change rate threshold at multiple adjacent time points, the aeolian sand area is determined to be shifting sand. Accumulated sand is judged based on the weighted gray value distribution and the rate of change of the weighted gray value in the aeolian sand area. The histogram of the weighted gray value in the aeolian sand area is obtained by calculation. If the histogram shows an obvious single peak and the peak gray value gradually increases, and the rate of change is less than the threshold value of the rate of change of mobile sand, the aeolian sand area is judged to be accumulated sand.

6. The method for identifying the change state of wind-blown sand on lightweight highway pavement according to claim 1 is characterized in that: The deformation characteristic factors include surface profile change factor and slope change factor; The surface profile variation factor calculation formula is: Among them, R edge is the surface profile variation factor, P1 and P2 are the edge profile perimeters of the surface profile obtained at time t1 and t2, respectively; The slope change factor calculation formula is: Among them, C flat is the surface flatness variation factor; σ1 and σ2 are the weighted gray value standard deviations of the aeolian sand area; The quantitative change characteristic factors include area change factor and gray value sum change factor; The area change factor calculation formula is: ΔA=A2-A1; Among them, ΔA is the area change factor, A1 and A2 are the areas of aeolian sand obtained at time t1 and t2 respectively; The calculation formula of the total gray value change factor is: ΔI S =I S2 -I S1 ; Among them, ΔI S is the total gray value change factor, I S1 and I S2 are the weighted gray value sums of all pixels in the aeolian sand area at different times.

7. The method for identifying the change state of wind-blown sand on lightweight highway pavement according to claim 1 is characterized in that: The calculation formula for the correlation coefficient between the deformation characteristic factor and quantitative characteristic factor of the aeolian sand and the meteorological data is: Among them, r ab is the correlation coefficient between a and b, a is a certain meteorological data, b is one of the deformation characteristic factors or one of the quantitative characteristic factors of aeolian sand, a i and b i are the values ​​of a and b in the i-th sample, and are the means of a and b respectively, and n is the number of samples.

8. The method for identifying the change state of wind-blown sand on lightweight highway pavement according to claim 1 is characterized in that: The aeolian sand deformation identification model is: Where D is the aeolian sand shape variable, r i is the correlation coefficient between meteorological factors and deformation characteristic factors, F 形变,i is the corresponding deformation characteristic factor, m is the number of meteorological factors involved in modeling; The aeolian sand quantitative change identification model is: Among them, Q is the variable of wind-blown sand, r j is the correlation coefficient between meteorological factors and quantitative characteristic factors, F 量变,j is the corresponding quantitative characteristic factor, and n is the number of meteorological factors involved in modeling.

9. The method for identifying the change state of wind-blown sand on lightweight highway pavement according to claim 1, characterized in that: The final aeolian sand change state determination method includes: Obtain the change state index of aeolian sand, that is, S=ω D ×D+ω Q ×Q; Among them, S is the index of wind-blown sand change state, ω D and ω Q are the weights of the aeolian sand deformation identification model and the weights of the aeolian sand quantitative change identification model, respectively. D and Q are the deformation and quantitative variables of aeolian sand, respectively. Preset the threshold range of wind-blown sand change state [S min ,S max ]; Determine the state of aeolian sand based on the threshold range; Among them, when S≤S min When , it is determined that the aeolian sand is in a stable state; When S min <S≤S max , aeolian sand is in a slowly changing state; When S>S max , aeolian sand is in a state of rapid change, and the state type of aeolian sand should be re-evaluated.

10. Lightweight highway pavement wind-blown sand change state recognition system, characterized by: The data acquisition module is used to use a drone equipped with a high-resolution camera to acquire image data along the highway and collect local meteorological data; The module for wind-blown sand area identification, weighted gray value calculation and state classification is used to calculate the pixel gray value of the image data, identify the wind-blown sand area, and obtain the weighted gray value of the wind-blown sand area on the road surface. By judging the weighted gray value, the wind-blown sand state is divided into static sand, mobile sand, and accumulated sand. The characteristic factor analysis and correlation coefficient calculation module is used to analyze the deformation characteristic factors and quantitative characteristic factors of aeolian sand in different aeolian sand states, and calculate the correlation coefficient between the deformation characteristic factors and quantitative characteristic factors of aeolian sand and meteorological data by using correlation coefficient analysis; The identification model establishment module is used to establish a deformation identification model of aeolian sand according to the deformation characteristic factors of aeolian sand and the correlation between the deformation characteristic factors of aeolian sand and meteorological data; and to establish a quantitative change identification model of aeolian sand according to the quantitative change characteristic factors of aeolian sand and the correlation between the quantitative change characteristic factors of aeolian sand and meteorological data; The wind-blown sand change state determination module is used to determine the weights of the wind-blown sand deformation recognition model and the wind-blown sand quantity change recognition model, and determine the final wind-blown sand change state in combination with the output values ​​of the wind-blown sand deformation recognition model and the wind-blown sand quantity change recognition model.

Images