Asphalt mixture aggregate uniformity analysis method and system based on Thiessen polygon distribution law

Abstract

The invention provides an asphalt mixture aggregate uniformity analysis method and system based on a Thiessen polygon distribution rule. According to the asphalt mixture aggregate uniformity analysis method and system based on the Thiessen polygon distribution rule, the asphalt aggregate uniformity is analyzed, the aggregate distribution uniformity can be accurately quantified, and a local non-uniform area is detected; and a scientific basis is provided for optimization of the mix proportion of the asphalt mixture and adjustment of a construction process. Compared with a traditional manual detection method, the detection efficiency is improved, the detection precision is improved, subjectivity and errors are effectively reduced, and the high-precision and high-efficiency requirements of engineering practice on aggregate uniformity detection are met.

Description

TECHNICAL FIELD

[0001] The application provides an asphalt mixture aggregate uniformity analysis method and system based on a Voronoi polygon distribution rule, and relates to the field of image data preprocessing. BACKGROUND

[0002] In road engineering, as the core pavement material, the aggregate distribution uniformity of asphalt mixture directly determines the mechanical properties and service life of the pavement. The traditional detection method relies on manual screening or two-dimensional porosity calculation, which has problems such as low efficiency, strong subjectivity, insufficient microscopic precision, and difficulty in capturing local differences in aggregate spatial distribution (Highway Asphalt Pavement Construction Technical Specification JTG F40-2020). With the development of digital image technology, quantitative analysis methods based on image processing have gradually emerged, promoting the automation and refinement of aggregate uniformity detection.

[0003] Existing image processing technologies (such as ImageJ and MATLAB image processing toolbox) mainly evaluate aggregate distribution through geometric feature indicators, such as:

[0004] Moment of inertia: By calculating the second moment feature of aggregate particles, it reflects the spatial distribution orientation and dispersion degree of particles, but it is insufficient in analyzing the adjacency relationship between multiple particles;

[0005] Area ratio and perimeter-area ratio: Statistics of aggregate total area ratio and particle shape compactness are used to preliminarily judge the aggregate density, but cannot quantify the uniformity of spatial distribution;

[0006] Grey level co-occurrence matrix: analyzes the spatial correlation of image gray level, and evaluates the interface distribution of aggregate and asphalt mastic, but has high computational complexity and depends on the accuracy of threshold segmentation;

[0007] Fractal dimension: describes the complexity of aggregate distribution through fractal theory, which is suitable for macroscopic uniformity evaluation, but has low sensitivity to local aggregation phenomenon.

[0008] However, these methods generally remain at the level of single-particle features or local area statistics, lacking global modeling of aggregate spatial position relationships. As a classic method of spatial division of discrete points, Voronoi polygon (Voronoi diagram) reflects the average distance between points through its polygon area, and the spatial symmetry through the number of adjacent edges, which can quantify the density and uniformity of aggregate distribution from multiple dimensions such as geometric shape (area, perimeter) and topological structure (edge number distribution). Currently, there is no technical solution to apply the area distribution rule and adjacency features of Voronoi polygon to asphalt aggregate uniformity analysis, and it is urgent to build a new quantitative evaluation system combining its advantages.

[0009] The aggregate segregation problem of asphalt mixture still lacks efficient detection means, the traditional method can only detect the overall gradation deviation, and the inertia moment and area ratio based on image cannot capture the mesoscopic difference of aggregate spatial distribution, resulting in lack of accurate basis for mixture optimization and construction regulation. SUMMARY

[0010] Therefore, in order to make up for the blank and deficiency of the prior art, the present application provides an asphalt mixture aggregate uniformity analysis method and system based on the Thiessen polygon distribution rule, which fills the blank of the prior art in the field of spatial distribution uniformity quantitative evaluation.

[0011] An asphalt mixture aggregate uniformity analysis method based on the Thiessen polygon distribution rule, characterized in that it comprises the following steps:

[0012] Step S1: image acquisition and preprocessing of the asphalt mixture test piece, including obtaining the original image of the asphalt mixture test piece and preprocessing it into a binary image, separating the aggregate and background of the binary image;

[0013] Step S2: using a method based on morphological processing and contour detection to identify aggregate particles in the preprocessed binary image. By calculating the centroid coordinates of each aggregate particle, its position in the image is determined;

[0014] Step S3: constructing the Thiessen polygon of the aggregate particles;

[0015] Step S4: uniformity index calculation of the Thiessen polygon;

[0016] Step S5: visualization and report generation of the results of the Thiessen polygon.

[0017] Further, step S1 includes the following content:

[0018] Step S11: using an industrial camera with a resolution greater than 5 million pixels to take pictures of the transverse and longitudinal sections of the asphalt mixture test piece, with light intensity controlled at 300-500 lux, to obtain the original image of the asphalt mixture test piece;

[0019] Step S12: grayscale the original image of the asphalt mixture test piece, including converting the original image of the asphalt mixture test piece into an 8-bit grayscale image using the weighted average method, the expression is:

[0020] Gray(x,y)=0.299·R(x,y)+0.587·G(x,y)+0.144·B(x,y)

[0021] Wherein, B(x, y) is the three-channel value of the color image at pixel point (x, y), and Gray(x, y) is the gray value, with a value range of [0, 255];

[0022] Step S13: removing noise of the gray image using a 3*3 or 5*5 pixel window, including the following contents:

[0023] f(x, y) = median{g(s, t) | (s, t) e W}

[0024] Wherein, W is a rectangular window centered at pixel point (x, y), g(s, t) is the original pixel value in the window, and median{} represents the median filter function.

[0025] Further, step S1 further includes the following contents:

[0026] Step S14: binarizing the gray image of the asphalt mixture specimen after removing noise by using Otsu automatic threshold method combined with manual adjustment, including the following contents:

[0027] Step S141: automatic threshold calculation: calculating the initial threshold T_0 by Otsu algorithm, which maximizes the inter-class variance of foreground and background based on the gray histogram, and the formula is:

[0028] T0 = argmax T [ω1(T)·ω2(T)·(μ1(T)-μ2(T)) 2 ]

[0029] Wherein, ω1, ω2 are the proportions of foreground and background pixels, and μ1, μ2 are the average gray values of foreground and background;

[0030] Step S142: manual adjustment: adjusting the threshold to 80≤T≤120 according to the image contrast, and the final binarization formula is:

[0031]

[0032] Wherein, b(x, y) is the pixel value after binarization,

[0033] When Gray(x, y)≥T, b(x, y) represents the asphalt mixture aggregate, which is considered as b(x, y)=1;

[0034] When Gray(x, y)<T, b(x, y) represents the asphalt mortar, which is considered as b(x, y)=0;

[0035] Through this step, the precise separation of aggregate and background is realized.

[0036] Further, step S2 includes the following contents:

[0037] Step S21: In the computer software, load the reference image containing the ruler, click the menu bar, enter the ruler setting interface, and establish the ruler parameters;

[0038] Step S22: Create a new ruler, set the unit, including the following processes:

[0039] Click the image, drag the green virtual ruler to the ruler position in the image, adjust the length to be consistent with the actual ruler, input the actual length value, click and apply the ruler to save the ruler configuration;

[0040] Step S23: Open the aggregate image, adjust the threshold to separate the aggregate from the background; use erosion and inflation to remove noise, fill holes, and separate adhered particles;

[0041] When the particles are adhered, use manual segmentation or automatic separation; ensure that each aggregate is an independent region to avoid centroid calculation errors;

[0042] Step S24: Use the software to automatically calculate the centroid coordinates of each aggregate, with the unit being the actual size set by the ruler; further, save the data of each aggregate, including the number, coordinates, and area information of each particle.

[0043] Further, step S3 includes the following content:

[0044] Step S31: For the identified aggregate particles, calculate their centroid coordinates (X i ,Y i ), and normalize them to the [0,1] interval by the following formula:

[0045]

[0046] Where X min , Y min are the coordinates of the top left corner of the image, W and H are the actual width and height of the image;

[0047] Where the normalization process eliminates the influence of image size on geometric analysis, making the area and number of edges of the Voronoi polygon have cross-scale comparability.

[0048] Further, step S3 also includes the following content:

[0049] Step S32: Import the normalized centroid coordinate matrix P = {(x' i , y' i )|i = 1, 2…, N} into the computer software, and call the voronoi function to generate the Voronoi polygon, where the function input and output relationship is:

[0050] [[vertices, regions] = voronoin(P)

[0051] wherein:

[0052] vertices is a matrix of polygon vertex coordinates (each row corresponds to the x', y' coordinates of a vertex);

[0053] regions is a matrix of region indices, each element corresponds to a list of polygon vertex indices of a centroid point.

[0054] Invalid region filtering: according to the actual boundary rectangle of the test piece [X min , X max ] × [Y min , Y max ], the vertices of the Thiessen polygon are denormalized to the actual coordinate system: x = x' · W + X min , y = y' · H + Y min ; and the Thiessen polygon completely contained in the boundary is retained, ensuring that the number of valid Thiessen polygons N valid ≥ 80, obtaining the Thiessen polygon set.

[0055] Further, step S4 includes the following content:

[0056] Step S41: Calculate the uniformity index of all Thiessen polygons in the Thiessen polygon set, including the following content:

[0057] Step S411: Calculate the average area of the Thiessen polygon set, including the average value of all valid cell areas,

[0058] Step S412: Calculate the area coefficient of variation of the Thiessen polygon set, which is the ratio of the standard deviation to the average value, expressed as follows:

[0059] where μ A is the mean line, and σ A is the standard deviation line;

[0060] Step S413: Calculate the ratio of the maximum area to the minimum area of the valid cells in the Thiessen polygon set;

[0061] Step S414: Calculate the circularity of the Thiessen polygon set;

[0062] Step S415: Calculate the average circularity of the valid cells in the Thiessen polygon set;

[0063] Step S416: Calculate the edge number variance of the Thiessen polygon set.

[0064] Further, step S5 includes the following content:

[0065] Step S51: Use the drawing function of computer software to display the distribution of Thiessen polygons in graphical form, and color code different areas according to the uniformity index;

[0066] Step S52: Color the Thiessen polygons according to the area coefficient of variation CVA value;

[0067] Step S53: Draw the Thiessen polygon area frequency distribution, superimpose the mean line and standard deviation line, and generate an area histogram;

[0068] Step S54: Generate the final asphalt mixture aggregate uniformity analysis report.

[0069] According to the second aspect of the present application, an asphalt mixture aggregate uniformity analysis system based on Thiessen polygon distribution law comprises an electronic device, wherein the electronic device comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the computer program to realize the asphalt mixture aggregate uniformity analysis method based on Thiessen polygon distribution law according to any one of the present application.

[0070] According to the third aspect of the present application, an asphalt mixture aggregate uniformity analysis system based on Thiessen polygon distribution law comprises a computer readable storage medium, and the computer readable storage medium stores a computer program, characterized in that the computer program is executed by a processor to realize the asphalt mixture aggregate uniformity analysis method based on Thiessen polygon distribution law according to any one of the present application.

[0071] The present application has the following advantages:

[0072] 1) High precision quantitative evaluation

[0073] By constructing Thiessen polygons, the present method can convert the spatial distribution of asphalt aggregates into specific polygon area characteristics. Through statistical analysis of these characteristics, the distribution uniformity of asphalt aggregates can be accurately quantified, avoiding the subjectivity and uncertainty of traditional methods.

[0074] 2) Comprehensive spatial analysis

[0075] Thiessen polygons can comprehensively consider the spatial position relationship between aggregates, not only can evaluate the overall distribution uniformity, but also can detect local aggregate aggregation or dispersion area. This helps to deeply understand the structural characteristics of asphalt mixture, and provides more accurate basis for optimizing the mix proportion and construction technology.

[0076] 3) High efficiency and automation

[0077] The MATLAB program is used to realize the automation of the whole analysis process, from image acquisition, preprocessing to the generation and analysis of the Voronoi polygon, which greatly improves the detection efficiency and reduces the time and error of manual operation. At the same time, the rich function library and powerful computing ability of MATLAB ensure the accuracy and reliability of the analysis results.

[0078] 4) Visualization of results

[0079] The method can display the distribution of the Voronoi polygon and the evaluation results of the aggregate distribution uniformity in the form of intuitive graphs and charts, such as the Voronoi diagram with color coding of different uniformity areas, area distribution histogram, etc. This visualization result facilitates the quick understanding and analysis of the aggregate distribution by engineering and technical personnel, helps to find problems in time and take corresponding improvement measures, and improves the efficiency of engineering decision-making. BRIEF DESCRIPTION OF DRAWINGS

[0080] Figure 1 The flowchart of the present application.

[0081] Figure 2 The Image Pro Plus software of the present application identifies and extracts the aggregate diagram.

[0082] Figure 3 The MATLAB function call relationship diagram for generating the Voronoi polygon of the present application.

[0083] Figure 4 The uniformity index calculation visualization result diagram of the present application.

[0084] Figure 5 The Voronoi diagram and area distribution histogram of the present application. DETAILED DESCRIPTION

[0085] The technical solutions of the present application will be described in detail below with reference to the accompanying drawings.

[0086] It should be noted that the following detailed description is exemplary and is intended to provide further explanation of the present application. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as generally understood by those skilled in the art to which the present application belongs.

[0087] It should be noted that the terms used herein are only for the purpose of describing the specific embodiments and are not intended to limit the exemplary embodiments according to the present application; as used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form, and furthermore, it should be understood that when the terms "comprise" and / or "include" are used in the specification, they mean the presence of a feature, step, operation, device, component and / or combination thereof.

[0088] As Figure 1 The flow chart of the present application shows that the present application proposes an asphalt mixture aggregate uniformity analysis method and system based on the Thiessen polygon distribution rule, wherein the asphalt mixture aggregate uniformity analysis method based on the Thiessen polygon distribution rule is realized based on the MATLAB program, and includes the following contents:

[0089] An asphalt mixture aggregate uniformity analysis method based on the Thiessen polygon distribution rule, characterized by comprising the following steps:

[0090] Step S1: image acquisition and preprocessing of the asphalt mixture test piece, including obtaining the original image of the asphalt mixture test piece and preprocessing it into a binary image, and separating the aggregate and the background of the binary image;

[0091] Step S2: using a method based on morphological processing and contour detection, the aggregate particles in the preprocessed binary image are identified. By calculating the centroid coordinates of each aggregate particle, its position in the image is determined;

[0092] Step S3: constructing the Thiessen polygon of the aggregate particles;

[0093] Step S4: calculating the uniformity index of the Thiessen polygon;

[0094] Step S5: visualization of the results of the Thiessen polygon and report generation.

[0095] In an embodiment of the present application, step S1 includes the following contents:

[0096] Step S11: using an industrial camera with a resolution greater than 5 million pixels, the cross section and the longitudinal section of the asphalt mixture test piece are photographed, the light intensity is controlled at 300-500 lux, and the original image of the asphalt mixture test piece is obtained;

[0097] Step S12: grayscale processing of the original image of the asphalt mixture test piece, including converting the original image of the asphalt mixture test piece into an 8-bit grayscale image using the weighted average method, and the expression is:

[0098] Gray(x,y)=0.299·R(x,y)+0.587·G(x,y)+0.144·B(x,y)

[0099] Wherein, B(x,y) is the three-channel value of the color image at pixel point (x,y), and Gray(x,y) is the gray value, with a value range of [0,255];

[0100] Step S13: removing the noise of the grayscale image using a 3x3 or 5x5 pixel window, including the following contents:

[0101] f(x,y)=median{g(s,t)∣(s,t)∈W}

[0102] Where W is a rectangular window centered on the pixel point (x, y), g(s, t) is the original pixel value in the window, and median{} represents the median filter function.

[0103] In one embodiment of the present invention, step S1 further includes the following:

[0104] Step S14: Binarize the grayscale image of the asphalt mixture specimen after noise removal using the Otsu automatic threshold method combined with manual adjustment, including the following:

[0105] Step S141: Automatic threshold calculation: Calculate the initial threshold T_0 using the Otsu algorithm, which maximizes the inter-class variance between foreground and background based on the grayscale histogram. The formula is:

[0106] T0=argmax T [ω1(T)·ω2(T)·(μ1(T)-μ2(T)) 2 ]

[0107] Among them, ω1 and ω2 are the ratios of foreground and background pixels, μ1 and μ2 are the average grayscale values ​​of foreground and background;

[0108] Step S142: Manual adjustment: According to the image contrast, the threshold is adjusted to 80≤T≤120. The final binarization formula is:

[0109]

[0110] Among them, b(x,y) is the pixel value after binarization, and this step is used to achieve accurate separation of aggregate and background.

[0111] In one embodiment of the present invention, step S2 includes the following contents:

[0112] Step S21: In the computer software, load the reference image containing the ruler, click the menu bar to enter the ruler setting interface, and establish the ruler parameters;

[0113] Step S22: Create a new ruler and set the unit, including the following process:

[0114] Click on the image, drag the green virtual ruler to the ruler position in the image, and adjust the length to match the actual ruler; enter the actual length value, click and apply the ruler to save the ruler configuration;

[0115] Step S23: open the aggregate image, adjust the threshold to separate the aggregate from the background; use erosion and dilation to remove noise, fill holes, and separate adhered particles;

[0116] When particles are adhered, use manual segmentation or automatic separation; ensure that each aggregate is an independent region to avoid centroid calculation errors;

[0117] Step S24: automatically calculate the centroid coordinates of each aggregate using software, with units of actual size set by the ruler; further, save the data of each aggregate, including the number, coordinates, and area information of each particle.

[0118] In an embodiment of the present application, step S3 includes the following content:

[0119] Step S31: for the identified aggregate particles, calculate their centroid coordinates (X i ,Y i ), and normalize them to the [0, 1] interval by the following formula:

[0120]

[0121] where X min , Y min are the coordinates of the top left corner of the image, W and H are the actual width and height of the image;

[0122] where the normalization process eliminates the influence of image size on geometric analysis, making the area and number of edges of the Voronoi polygon have cross-scale comparability.

[0123] In an embodiment of the present application, step S3 also includes the following content:

[0124] Step S32: import the normalized centroid coordinate matrix P = {(x' i , y' i )|i = 1, 2…, N} into computer software and call the voronoi function to generate the Voronoi polygon, where the function input-output relationship is:

[0125] [[vertices, regions] = voronoin(P)

[0126] where:

[0127] vertices is the polygon vertex coordinate matrix (each row corresponds to the x', y' coordinates of a vertex);

[0128] regions is the region index matrix, each element corresponds to a list of polygon vertex indices of the centroid point.

[0129] Invalid region filtering: according to the actual boundary rectangle of the specimen [X min , Xmax ]×[Y min ,Y max ], denormalize the Thiessen polygon vertices to the actual coordinate system: x = x'·W+X min ,,y=y'·H+Y min ; and retain the Thiessen polygons that are completely contained in the boundary to ensure the number of valid Thiessen polygons N valid ≥80, get the Thiessen polygon set.

[0130] In one embodiment of the present invention, step S4 includes the following contents:

[0131] Step S41: Calculate the uniformity index of all Thiessen polygons in the Thiessen polygon set, including the following:

[0132] Step S411: Calculate the average area of ​​the Thiessen polygon set, including the average area of ​​all valid units.

[0133] Step S412: Calculate the area variation coefficient of the ratio of the standard deviation of the area of ​​the Thiessen polygon set to the mean value, which is expressed as follows:

[0134] where μ A is the mean line, σ A is the standard deviation line;

[0135] Step S413: Calculate the ratio of the maximum area to the minimum area of ​​the valid cells in the Thiessen polygon set;

[0136] Step S414: Calculate the roundness of the Thiessen polygon set;

[0137] Step S415: Calculate the average circularity of the valid cells in the Thiessen polygon set;

[0138] Step S416: Calculate the variance of the number of edges in the Thiessen polygon set.

[0139] In one embodiment of the present invention, step S5 includes the following contents:

[0140] Step S51: using the drawing function of the computer software, the distribution of the Thiessen polygons is displayed in a graphical form, and different areas are color-coded according to the uniformity index;

[0141] Step S52: Coloring the Thiessen polygons according to the coefficient of variation of area (CVA) values;

[0142] Step S53: draw the frequency distribution of Thiessen polygon area, superimpose the mean line and standard deviation line, and generate an area histogram;

[0143] Step S54: generating a final asphalt mixture aggregate uniformity analysis report.

[0144] According to a second aspect of the present application, an asphalt mixture aggregate uniformity analysis system based on the Thiessen polygon distribution rule comprises an electronic device, wherein the electronic device comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, and wherein the processor implements the asphalt mixture aggregate uniformity analysis method based on the Thiessen polygon distribution rule according to any one of the embodiments of the present application when executing the computer program.

[0145] According to a third aspect of the present application, an asphalt mixture aggregate uniformity analysis system based on the Thiessen polygon distribution rule comprises a computer readable storage medium, wherein the computer readable storage medium stores a computer program, and wherein the computer program implements the asphalt mixture aggregate uniformity analysis method based on the Thiessen polygon distribution rule according to any one of the embodiments of the present application when executed by a processor.

[0146] In addition to the above, the present application also includes the following embodiments:

[0147] In an embodiment of the present application, step S142 comprises the following:

[0148] Manual fine tuning: according to the image contrast, the threshold value is adjusted to 80≤T≤120, and the final binarization formula is:

[0149]

[0150] wherein b(x, y) is the pixel value after binarization,

[0151] When Gray(x, y)≥T, b(x, y) represents asphalt mixture aggregate, and is regarded as b(x, y)=1;

[0152] When Gray(x, y)<T, b(x, y) represents asphalt mortar, and is regarded as b(x, y)=0;

[0153] Through this step, the accurate separation of aggregate and background is realized.

[0154] In an embodiment of the present application, step S2 is realized by using software Image-Pro Plus, and comprises the following:

[0155] Step S21: opening Image-Pro Plus and loading a reference image containing a ruler. Clicking the menu bar Measure>Calibration>Spatial enters the ruler setting interface to establish the ruler parameters.

[0156] Step S22: Click New to create a new ruler, set the unit (e.g. meter); click Image, drag the green virtual ruler to the ruler position in the image, adjust the length to be consistent with the actual ruler (e.g. 50 μm); input the actual length value (e.g. 50), click OK and apply the ruler (Apply), save the ruler configuration.

[0157] In the Spatial Calibration interface, name the ruler (e.g. "Aggregate_Cal") for subsequent calling

[0158] Step S23: Then image preprocessing and particle segmentation, binarization processing, including the following contents:

[0159] Open the aggregate image, click Edit>Convert to Binary, adjust the threshold to separate aggregate and background; use Morphology Tools (erode→dilate) to remove noise, fill holes, and separate adherent particles

[0160] If the particles are adherent, use Edit>Split Objects to manually segment, or through Count / Size>Separate Touching Objects to automatically separate; ensure that each aggregate is an independent region to avoid centroid calculation errors.

[0161] Step S24: Extract the centroid coordinates, automatically measure the centroid

[0162] Click Measure, check Center (centroid) and ObjectArea (area, used to filter small noise); the software automatically calculates the centroid coordinates (X, Y) of each aggregate, with the unit being the actual size set by the ruler. Export the data

[0163] In the File window, right-click to select Export Data, save as Excel, containing the number, X / Y coordinates, area, etc. of each particle. Finally, normalize the coordinates, where W and H are the actual width and height of the image (calculated by the ruler, e.g. image width 2000 μm). Operation: Add two columns in Excel, input the formula to calculate the normalized coordinates; save the normalized data for subsequent analysis.

[0164] In one embodiment of the present application, step S32 uses MATLAB to call the voronoi function to generate the Voronoi polygon.

[0165] In one embodiment of the present application, the meaning of the uniformity index calculation in step S4 is shown in Table 1:

[0166]

[0167]

[0168] Table 1 uniformity index summary table

[0169] In addition, the present application also provides an embodiment comprising the following contents:

[0170] Step S1: image acquisition: select a certain asphalt mixture test piece, use an industrial camera with a resolution of 2000x2000 pixels to collect its cross-sectional image, the actual size of the image is 20mmx20mm (i.e. W=20000um, H=20000um), then the normalized coordinates of the particles are:

[0171]

[0172] Step S2: preprocessing and aggregate recognition: in Image Pro Plus, the image is grayed, median filter denoising and binarization processing, and the morphological operation is used to segment the adherent aggregate particles, a total of 130 aggregate particles are identified, the centroid coordinates of each aggregate are extracted and normalized, and the results are shown in Figure 2 .

[0173] Step S3: Voronoi polygon generation: the normalized centroid coordinates are imported into MATLAB, and the voronoin function is used to generate Voronoi polygons, and the invalid area outside the boundary is removed to obtain 83 effective Voronoi polygons. The related code is shown in Figure 3 .

[0174] Step S4: uniformity index calculation: the area average is 0.0074, the CVA is 0.3015; the maximum / minimum area ratio is 5.0203, the shape regularity index is the average circularity 07861, the edge number analysis, the average edge number is 5.96, and the edge number variance is 1.3036.

[0175] Step S5: visualization and report: generate Voronoi diagram, draw area distribution histogram and mark mean line, generate analysis report containing index calculation results and visualization graphics, as shown in Figure 4 、 Figure 5 .

[0176] The above is the preferred embodiment of the present application, any changes made according to the technical solutions of the present application, as long as the generated function does not exceed the scope of the technical solutions of the present application, belongs to the protection scope of the present application.

Claims

1. A method for analyzing the uniformity of asphalt mixture aggregate based on the Thiessen polygon distribution law, characterized in that: It includes the following steps: Step S1: Image acquisition and preprocessing of asphalt mixture specimens, including obtaining the original image of the asphalt mixture specimen and preprocessing it into a binary image, and separating the aggregates and the background of the binary image; Step S2: Adopting a method based on morphological processing and contour detection to identify aggregate particles in the preprocessed binary image, and determining their positions in the image by calculating the centroid coordinates of each aggregate particle; Step S3: Constructing the Thiessen polygon of aggregate particles; Step S4: Calculating the uniformity index of the Thiessen polygon; Step S5: Visualization of the results of the Thiessen polygon and report generation.

2. The asphalt mixture aggregate uniformity analysis method based on Thiessen polygon distribution law according to claim 1, characterized in that: Step S1 includes the following contents: Step S11: Using an industrial camera with a resolution greater than 5 million pixels to photograph the cross-section and longitudinal section of the asphalt mixture specimen, controlling the light intensity at 300 - 500 lux, and obtaining the original image of the asphalt mixture specimen; Step S12: Grayscaling the original image of the asphalt mixture specimen obtained, including converting the original image of the asphalt mixture specimen into an 8-bit grayscale image by using the weighted average method, and the expression is: Gray(x,y) = 0.299·R(x,y) + 0.587·G(x,y) + 0.144·B(x,y) in, B(x,y) is the three-channel value of the color image at the pixel point (x,y), Gray(x,y) is the grayscale value, and the value range is [0,255]; Step S13: Removing the noise of the grayscale image by using a 3×3 or 5×5 pixel window, including the following contents: f(x,y) = median{g(s,t)∣(s,t)∈W} where W is a rectangular window centered on the pixel point (x,y), g(s,t) is the original pixel value within the window, and median{} represents the median filtering function.

3. The asphalt mixture aggregate uniformity analysis method based on Thiessen polygon distribution law according to claim 2, characterized in that: Step S1 also includes the following contents: Step S14: Binarizing the grayscale image of the asphalt mixture specimen after removing the noise by using the Otsu automatic threshold method combined with manual adjustment, including the following contents: Step S141: Automatic threshold calculation: Calculating the initial threshold T_0 through the Otsu algorithm, which is based on maximizing the between-class variance of the foreground and the background in the grayscale histogram, and the formula is: T0=argmax T [ω1(T)·ω2(T)·(μ1(T)-μ2(T)) 2 where ω1, ω2 are the ratios of foreground and background pixels, and μ1, μ2 are the average grayscale values of the foreground and the background; Step S142: Manual adjustment: Adjusting the threshold to 80 ≤ T ≤ 120 according to the image contrast, and the final binarization formula is: where b(x,y) is the pixel value after binarization, when Gray(x,y) ≥ T, b(x,y) represents the aggregates of the asphalt mixture and is regarded as b(x,y) = 1; Gray(x,y) < T, b(x,y) represents the asphalt mortar and is regarded as b(x,y) = 0; Through this step, the accurate separation of aggregates and the background is achieved.

4. The method for analyzing the uniformity of asphalt mixture aggregate based on Thiessen polygon distribution law according to claim 3, characterized in that: Step S2 includes the following contents: Step S21: In the computer software, load the reference image containing the scale, click on the menu bar, enter the scale setting interface, and establish the scale parameters; Step S22: Creating a new scale and setting the unit, including the following process: Click on the image, drag the green virtual scale to the scale position in the image, adjust the length to be consistent with the actual scale; input the actual length value, click and apply the scale to save the scale configuration; Step S23: Open the aggregate image, adjust the threshold to separate the aggregate from the background; use corrosion and expansion to remove noise, fill holes, and segment adhered particles; When particles are stuck together, use manual segmentation or automatic separation to ensure that each aggregate is an independent area to avoid centroid calculation errors; Step S24: using software to automatically calculate the centroid coordinates of each aggregate, with the unit being the actual size set by the ruler; further, saving the data of each aggregate including the number, coordinates, and area information of each particle.

5. The asphalt mixture aggregate uniformity analysis method based on Thiessen polygon distribution law according to claim 4 is characterized in that: Step S3 includes the following contents: Step S31: Calculate the centroid coordinates (X i ,Y i ), and normalized to the interval [0,1] using the following formula: Among them, X min 、Y min is the coordinate of the upper left corner of the image, W and H are the actual width and height of the image; Among them, the normalization process eliminates the influence of image size on geometric analysis, making the area and number of sides of Thiessen polygons comparable across scales.

6. The method for analyzing the uniformity of asphalt mixture aggregate based on Thiessen polygon distribution law according to claim 5, characterized in that: Step S3 also includes the following: Step S32: Normalize the centroid coordinate matrix P = {(x ' i ,y ' i )|i=1,2…,N} is imported into the computer software and the voronoi function is called to generate Thiessen polygons. The input and output relationship of the function is: [[vertices,regions]=voronoin(P) in: vertices is the polygon vertex coordinate matrix (each row corresponds to the x of a vertex ' ,y ' coordinate); regions is the region index matrix, each element corresponds to a polygon vertex index list of a centroid point; Invalid area filtering: According to the actual boundary rectangle of the specimen [X min, X max ]×[Y min, Y max ], denormalize the Thiessen polygon vertices to the actual coordinate system: x = x ' W+X min, ,y=y ' ·H+Y min ; and retain the Thiessen polygons that are completely contained in the boundary to ensure the number of valid Thiessen polygons N valid ≥80, get the Thiessen polygon set.

7. The method for analyzing the uniformity of asphalt mixture aggregate based on Thiessen polygon distribution law according to claim 6, characterized in that: Step S4 includes the following contents: Step S41: Calculate the uniformity index of all Thiessen polygons in the Thiessen polygon set, including the following: Step S411: Calculate the average area of ​​the Thiessen polygon set, including the average area of ​​all valid units. Step S412: Calculate the area variation coefficient of the ratio of the standard deviation of the area of ​​the Thiessen polygon set to the mean value, which is expressed as follows: where μ A is the mean line, σ A is the standard deviation line; Step S413: Calculate the ratio of the maximum area to the minimum area of ​​the valid cells in the Thiessen polygon set; Step S414: Calculate the roundness of the Thiessen polygon set; Step S415: Calculate the average circularity of the valid cells in the Thiessen polygon set; Step S416: Calculate the variance of the number of edges in the Thiessen polygon set.

8. The method for analyzing the uniformity of asphalt mixture aggregate based on Thiessen polygon distribution law according to claim 1, characterized in that: Step S5 includes the following contents: Step S51: using the drawing function of the computer software, the distribution of the Thiessen polygons is displayed in a graphical form, and different areas are color-coded according to the uniformity index; Step S52: Coloring the Thiessen polygons according to the coefficient of variation of area (CVA) values; Step S53: draw the frequency distribution of Thiessen polygon area, superimpose the mean line and standard deviation line, and generate an area histogram; Step S54: Generate a final asphalt mixture aggregate uniformity analysis report.

9. An asphalt mixture aggregate uniformity analysis system based on Thiessen polygon distribution law, comprising an electronic device, wherein the electronic device comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, it implements the asphalt mixture aggregate uniformity analysis method based on the Thiessen polygon distribution law as described in any one of claims 1 to 8.

10. An asphalt mixture aggregate uniformity analysis system based on Thiessen polygon distribution law, comprising a computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method for analyzing the uniformity of asphalt mixture aggregate based on the Thiessen polygon distribution law as described in any one of claims 1 to 8 is implemented.

Images