Specific quantification method for dispersion uniformity of modifier in asphalt based on image processing technology

Abstract

The invention relates to the technical field of road engineering, in particular to a method for specifically quantifying the dispersion uniformity of a modifier in asphalt based on an image processing technology, which comprises the following steps of: 1, preparing a polymer modified asphalt sample of which the observation effect meets the requirement, and obtaining a fluorescence microscopic true color image under the observation of a fluorescence microscope; 2, carrying out conversion, enhancement, segmentation and marking processing on the fluorescent microscopic true color image obtained in the step 1: sequentially carrying out gray level image conversion, adaptive histogram equalization, Otsus threshold segmentation and skeletonized image feature pixel point marking on the fluorescent microscopic true color image; 3, based on the Otsus threshold segmentation image and the skeletonized image obtained in the step 2, carrying out feature pixel point marking on the skeletonized image; the average swelling diameter ratio, the shape factor average value and the mass center standard deviation of the polymer modifier are calculated respectively, and the dispersion uniformity of the polymer modifier in the matrix asphalt is judged. According to the invention, the dispersion uniformity of the polymer modifier in asphalt can be represented from multiple aspects.

Description

Technical Field

[0001] The present invention relates to the technical field of road engineering, and in particular to a specific quantitative method for the dispersion uniformity of a modifier in asphalt based on image processing technology. Background Art

[0002] Asphalt is a viscoelastic material with time- and temperature-dependent properties, that is, at high temperatures, asphalt exhibits Newtonian fluid characteristics, at intermediate temperatures, it behaves as a viscoelastic liquid, and at low temperatures, it behaves like a brittle material. Therefore, in asphalt pavements, the viscosity of asphalt decreases due to the high temperature environment, and then the asphalt mixture produces rutting damage under the action of vehicle loads. At the same time, due to the stress relaxation cycle of vehicle loads, damage to the asphalt pavement accumulates, resulting in fatigue cracking. In addition, asphalt tends to a vitrified state in a low-temperature environment, its stress dispersion ability is weakened, and it is easy to crack. It can be concluded that most of the diseases of asphalt pavements are related to the rheology and durability of asphalt, and the change of asphalt performance with time and temperature has a significant impact on the occurrence of asphalt pavement diseases.

[0003] In order to solve or alleviate the above problems, researchers have tried to use different types of polymer modifiers to change the properties of asphalt, such as styrene-butadiene-styrene (SBS), styrene-butadiene rubber (SBR), polyurethane (PU), epoxy resin (EP), polyphosphoric acid (PPA), ethylene vinyl acetate (EVA), high-density polyethylene (HDPE), low-density polyethylene (LDPE), ethylene propylene diene terpolymer (EPDM) and reactive ethylene terpolymers (RET). Better polymer modifiers can improve one or more basic properties of asphalt, including rigidity, elasticity, brittleness, storage stability and durability. Therefore, from the perspective of pavement service life, choosing the right polymer modifier can effectively achieve the required performance of asphalt. However, due to the differences in the polarity, relative molecular mass and solubility parameters of polymer modifiers and asphalt, the polymer modifiers are easily unevenly dispersed in the asphalt during the preparation process, resulting in agglomeration, which in turn affects the storage stability of the polymer modified asphalt and eventually leads to stratification. In order to ensure that the performance of the produced polymer modified asphalt is excellent and stable, and to make the polymer modifier evenly dispersed in the asphalt, the fluorescence microscope test is one of the simplest and most effective detection methods.

[0004] Fluorescence microscopy is used to detect the dispersibility of polymer modifier particles in matrix asphalt. When the modifier particles are blended, swelled and chemically reacted in asphalt, they will combine with the aromatic hydrocarbon components in the asphalt, causing the modifier particles to emit fluorescence and show a light-colored modified phase in the dark asphalt phase. At present, there are two main methods for quantitatively evaluating the dispersibility of polymer modifiers in asphalt using fluorescence microscopy. One method uses MATLAB and Image Pro Plus software to analyze the area ratio of modifier fluorescent pixels to total pixels to evaluate the uniformity of modifier distribution. This method focuses on the area ratio of modifier to asphalt, characterizing the limited uniformity of modifier dispersion. The other method uses MATLAB and AutoCAD software to binarize and skeletonize the fluorescent image, and calculates the ratio of the area (number of white pixels in the binarized image) and the centerline length (number of white pixels in the skeletonized image) of the polymer modifier to evaluate its dispersion uniformity. However, this method assumes that the shape of the polymer modifier is regular, such as round or square. In the fluorescence microscopy image, the shape of the polymer modifier may be irregular. In addition, it is assumed that the dispersed shape of the polymer modifier is a regular circle or square, but their dispersed area can be large or small, which cannot well characterize the uniformity of the dispersion of the polymer modifier. Secondly, this method only considers the ratio of the area of ​​the polymer modifier to the center line length, and does not consider the position of the polymer modifier in the fluorescence microscopy image. If the polymer modifier is mainly concentrated in a part of the fluorescence microscopy image and less in other areas, it is also difficult to evaluate the uniformity of the dispersion of the polymer modifier. Summary of the invention

[0005] In view of the shortcomings of the prior art, the purpose of the present invention is to provide a specific quantitative method for the dispersion uniformity of the modifier in asphalt based on image processing technology.

[0006] To achieve the above object, the present invention provides the following technical solution: a specific quantitative method for the dispersion uniformity of a modifier in asphalt based on image processing technology, the steps of which are:

[0007] Step 1: Prepare a polymer-modified asphalt sample that meets the observation requirements, and obtain a fluorescence microscopic true color image under the observation of a fluorescence microscope;

[0008] Step 2: convert, enhance, segment and mark the fluorescence microscopy true color image obtained in step 1: perform grayscale image conversion, adaptive histogram equalization, Otsu's threshold segmentation, and skeleton image feature pixel point marking on the fluorescence microscopy true color image in sequence;

[0009] Step 3: Based on the Otsu's threshold segmentation image and skeletonized image obtained in step 2, the average swelling diameter ratio, shape factor average value and centroid standard deviation of the polymer modifier are calculated respectively to determine the dispersion uniformity of the polymer modifier in the matrix asphalt.

[0010] In some of the embodiments, in step 2, grayscale image conversion, adaptive histogram equalization, Otsu's threshold segmentation, and skeletonized image feature pixel marking are sequentially performed on the fluorescence microscopy image with the assistance of Python code.

[0011] In some of the embodiments, the average swelling diameter ratio, shape factor average value and centroid standard deviation of the polymer modifier in step 3 are implemented with the assistance of Python code.

[0012] In some embodiments, according to step 3,

[0013] The average swelling diameter ratio μD is calculated using the following formula:

[0014] Among them, T o is the total number of main object pixels, i.e., the total number of polymer modifiers represented by the total number of white pixels; T s is the total number of skeletonized pixels, that is, the marked modifier pixels are processed by the skeletonization algorithm to reduce them to a "skeleton" with a width of one pixel while maintaining the overall structure and connectivity of the modifier.

[0015] In some embodiments, according to step 3,

[0016] The average shape factor μS is calculated using the following formula:

[0017] Wherein, A is the area of ​​the white area in the segmented image, that is, the area of ​​the polymer modifier; P is the perimeter of the white area in the segmented image, that is, the perimeter of the polymer modifier.

[0018] In some embodiments, according to step 3,

[0019] The centroid standard deviation includes the centroid X-axis standard deviation and the standard deviation of the Y axis of the centroid

[0020] The following formulas are used for calculation:

[0021]

[0022] Among them, x i and i are the x- and y-coordinates of the centroid of the connected region of the ith polymer modifier, and are the average x- and y-coordinates of the centroids of all polymer modifier connected regions, respectively. N is the total number of polymer modifier connected regions, and the centroid is the center position of a modifier connected region.

[0023] Compared with the prior art, the beneficial effects of the present invention are: the average swelling diameter ratio, shape factor average value and centroid standard deviation of the polymer modifier in the fluorescent microscopic image are calculated, and a comprehensive and accurate method for calculating the characteristics of the polymer modifier can be provided, and the dispersion uniformity of the polymer modifier in the asphalt can be characterized from multiple aspects, and the modification effect of the polymer modified asphalt can be judged, and a quantitative basis can be provided for the preparation process effect of the modified asphalt. In addition, in the process of calculating the dispersion uniformity, the present invention only needs to be able to run the Python code environment, and there is no need to pay extra to purchase copyright software for auxiliary calculation. Since all steps are in the form of code, there is no need to learn additional software operation methods. Only the code needs to be simply run to realize the calculation of the dispersion uniformity of the polymer modifier in the fluorescent microscopic image, and the fluorescent microscopic images can be processed in batches, which is more efficient and fast, and has good practical value and promotion value.

[0024] Details of one or more embodiments of the present application are presented in the following drawings and descriptions to make other features, purposes and advantages of the present application more concise and easy to understand. The present application is fully described and understood through the embodiments of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 Schematic diagram of the process of the present invention;

[0026] Figure 2 This is a sample of fluorescence microscope images of polymer modified asphalt;

[0027] Figure 3 This is a sample image of a fluorescence microscopy image after grayscale processing;

[0028] Figure 4 This is a sample image of a fluorescence microscopy image after adaptive histogram equalization processing;

[0029] Figure 5 This is a sample image of a fluorescence microscopy image after Otsu's threshold segmentation processing;

[0030] Figure 6 This is a sample image of a fluorescence microscopy image after the skeletonized image feature pixel points are marked;

[0031] Figure 7The images of the method of the present invention are obtained by using a sample of polymer-modified asphalt with a polymer modifier dosage of 5%, wherein (a), (b), (c) and (d) are respectively the true color fluorescence microscopy image, the adaptive histogram equalization image, the Otsu's threshold segmentation image, and the skeletonized image feature pixel point marking image in the embodiment;

[0032] Figure 8 The images of the method of the present invention are obtained by using a sample of polymer-modified asphalt with a polymer modifier dosage of 10%, wherein (a), (b), (c) and (d) are respectively the true color fluorescence microscopy image, the adaptive histogram equalization image, the Otsu's threshold segmentation image, and the skeletonized image feature pixel point marking image in the embodiment;

[0033] Fig. 9 The images of the method of the present invention are obtained by using a sample of polymer-modified asphalt with a polymer modifier dosage of 15%, wherein (a), (b), (c) and (d) are respectively the true color fluorescence microscopy image, the adaptive histogram equalization image, the Otsu's threshold segmentation image, and the skeletonized image feature pixel point marking image in the embodiment;

[0034] Fig.10 The images involved in the method of the present invention for a sample of polymer modified asphalt with a polymer modifier dosage of 20%, wherein (a), (b), (c) and (d) are respectively the fluorescence microscopy true color image, the adaptive histogram equalization image, the Otsu's threshold segmentation processed image, and the skeletonized image feature pixel point marking image in this embodiment. DETAILED DESCRIPTION

[0035] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0036] See also Figure 1 The present invention provides a technical solution: a specific quantitative method for the uniformity of dispersion of a modifier in asphalt based on image processing technology, the steps of which are:

[0037] Step 1: Prepare a polymer modified asphalt sample that meets the observation requirements, and obtain a fluorescence microscopic true color image (such as Figure 2 Samples are shown for processing);

[0038] Step 2: convert, enhance, segment and mark the fluorescence microscopy true color image obtained in step 1: perform grayscale image conversion, adaptive histogram equalization, Otsu's threshold segmentation, and skeleton image feature pixel point marking on the fluorescence microscopy true color image in sequence;

[0039] Step 3: Based on the Otsu's threshold segmentation image and skeletonized image obtained in step 2, the average swelling diameter ratio, shape factor average value and centroid standard deviation of the polymer modifier are calculated respectively to determine the dispersion uniformity of the polymer modifier in the matrix asphalt.

[0040] In step 2, the fluorescence microscopy image is successively converted to grayscale, subjected to adaptive histogram equalization, processed by Otsu's threshold segmentation, and skeletonized image feature pixel marking, which are implemented with the assistance of Python code.

[0041] In step 3, the average swelling diameter ratio, shape factor average value and centroid standard deviation of the polymer modifier are implemented with the assistance of Python code.

[0042] According to step 3,

[0043] The mean swelling diameter ratio μD (Mean Swelling Diameter Ratio, μD) is calculated using the following formula:

[0044] Among them, T o Total Count of Object Pixels, T o ), i.e., the total number of white pixels represents the total number of polymer modifiers; T s is the total number of skeletonized pixels (Total Count of SkeletonizedPixels, T s ), that is, the marked modifier pixels are processed by a skeletonization algorithm to reduce them to a "skeleton" of single pixel width while maintaining the overall structure and connectivity of the modifier.

[0045] The mean shape factor μS (Mean Shape Factor, μS) is calculated using the following formula:

[0046] Wherein, A is the area of ​​the white area in the segmented image, that is, the area of ​​the polymer modifier; P is the perimeter of the white area in the segmented image, that is, the perimeter of the polymer modifier.

[0047] The centroid standard deviation includes the centroid X-axis standard deviation (Standard Deviation of Centroid X-Coordinates, ) and the Y-axis standard deviation of the centroid (Standard Deviation ofCentroid Y-Coordinates, ),

[0048] The following formulas are used for calculation:

[0049]

[0050] Among them, x i and i are the x- and y-coordinates of the centroid of the connected region of the ith polymer modifier, and are the average x- and y-coordinates of the centroids of all polymer modifier connected regions, respectively. N is the total number of polymer modifier connected regions, and the centroid is the center position of a modifier connected region.

[0051] Through this technical solution, the processing process only uses Python code, and the environment that can run the code can be processed. The processing method is as follows:

[0052] Step 1: Load the image file to be processed through img = Image.open (file_path). The opened image file can be various common image formats, such as JPEG, PNG, BMP, etc. In this embodiment, a true color image of a fluorescence microscope is loaded; then, through img_gray = img.convert ('L'), the convert method of the image processing library (PIL library) is used to convert the opened true color image into a grayscale image, so that the values ​​of the red, green, and blue channels of each pixel are converted into a grayscale value. Usually, this grayscale value is obtained by taking a weighted average of the values ​​of the red, green, and blue channels; finally, the grayscale image is saved through img_gray.save (output_path, dpi = (dpi, dpi)), and the converted grayscale image can be saved to the specified output path (output_path). When saving a grayscale image, you can set the DPI (dots per inch) of the image to control the resolution of the image, and get the following Figure 3 Shown is a sample fluorescence microscopy image after grayscale processing.

[0053] Step 2: Through clahe = cv2.createCLAHE (clipLimit = clip_limit, tileGridSize = tile_grid_size), Figure 3 Grayscale the sample image to create a contrast limited adaptive histogram equalization (CLAHE) object. CLAHE is a technique for improving image contrast. In this command, two parameters can be set: clipLimit and tileGridSize. The clipLimit parameter is used to control the maximum contrast enhancement. When the clipLimit value is larger, the contrast enhancement is greater and the image details are clearer. The tileGridSize parameter is used to control the size of the area for histogram equalization. When the tileGridSize value is smaller, the contrast in each small area will be enhanced independently, which can better preserve the local details of the image. In this embodiment, clip_limit=2.0 and tile_grid_size=(8,8) are used for control so that Figure 3 Grayscale sample images are enhanced to improve contrast and make image details clearer. Figure 4 A sample of a fluorescence microscopy image after adaptive histogram equalization is shown in the figure.

[0054] Step 3: Use the threshold function of the OpenCV library to input img_otsu=cv2.threshold(img,0,255,cv2.THRESH_BINARY+cv2.THRESH_OTSU). Figure 4 The adaptive histogram equalization process performs Otsu's threshold processing. The second and third parameters of this function are set to 0 and 255 respectively, indicating that Figure 4 The grayscale value is divided into two intervals: [0, threshold) and [threshold, 255]. The pixels in these two intervals will be set to 0 (black) and 255 (white) respectively. Figure 4 The image is divided into foreground and background, which is a binary image in which pixels have only two possible values: 0 (black) and 255 (white). Figure 5 A sample of a fluorescence microscopy image after Otsu's threshold segmentation is shown in the figure.

[0055] Step 4: Use skeleton=skeletonize(binary_img) to Figure 5 The Otsu's threshold segmentation processing graph is summed to obtain the number of white pixels (X), which is the number of polymer modifier pixels. Figure 5 The skeletonization process is performed to obtain a sample image of the fluorescence microscopy image after the skeletonized image feature pixel point marking process, such as Figure 6 As shown. Figure 6 Perform a sum operation to obtain the number of white pixels (Y) in the skeletonized image.

[0056] Step 5: Calculate the average swelling diameter ratio by diameter = X / Y if Y! = 0 else 0, where X is Figure 5 Otsu's threshold segmentation process the number of white pixels in the image; Y is Figure 6 The number of white pixels after the skeletonized image feature pixel marking process.

[0057] Step 6: Calculate the mean shape factor (μS) by using mean_shape_factor = np.mean(shape_factors). The formula for this command is: Among them, A is Figure 5 The area of ​​the white area in the Otsu's threshold segmentation processing diagram is the area of ​​the polymer modifier; P is Figure 5 The perimeter of the white area in the Otsu's threshold segmentation process is the perimeter of the polymer modifier.

[0058] Step 7: Calculate the Standard Deviation of Centroid X-Coordinates, by std_centroids_x = np.std(centroids_x) and std_centroids_y = np.std(centroids_y). ) and Standard Deviation of Centroid Y-Coordinates, ). The formula for this command is: and Among them, x i and i are the x- and y-coordinates of the centroid of the connected region of the ith polymer modifier, and are the average x- and y-coordinates of the centroids of all polymer modifier connected regions, respectively, and N is the total number of polymer modifier connected regions. The centroid is the central position of a modifier connected region.

[0059] Step 8: Through results = pd.DataFrame(columns = ['Image','Number of white pixels','Number of skeleton pixels','Relative swelling diameter','Mean shapefactor','Std of centroids x','Std of centroids y']), all the calculated data in steps 4 to 8 are stored in the results.csv file to obtain the data shown in Table 1.

[0060] Table 1

[0061]

[0062] Similarly, the fluorescence microscopy images of polymer modified asphalt with polymer modifier dosage of 5%, 10%, 15% and 20% were collected, as shown in Figure 2. Figures 7-10 The average swelling diameter ratio, shape factor average value and centroid standard deviation of the polymer modifier were calculated using the method of the present invention for each sample image. The process was processed by computer batch processing, and the calculation results are shown in Table 2.

[0063] Table 2

[0064]

[0065] From the calculation results of the above processing of multiple samples, it can be seen that the specific quantitative method of the dispersion uniformity of the polymer modifier in the matrix asphalt based on the image processing technology is used to calculate the average swelling diameter ratio, shape factor average value and centroid standard deviation of the polymer modifier, taking into account the morphology and position of the polymer modifier in the asphalt, and more comprehensively, accurately and quantitatively evaluating the dispersion uniformity of the polymer modifier in the asphalt. In addition, in the process of calculating the dispersion uniformity, the present invention only needs to be able to run the Python code environment, without learning the software operation method, and can process pictures in batches, which is more efficient and fast, and easy to promote.

[0066] The above embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

[0067] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A specific quantitative method for the uniformity of dispersion of a modifier in asphalt based on image processing technology, characterized in that: The steps are: Step 1: Prepare a polymer-modified asphalt sample that meets the observation requirements, and obtain a fluorescence microscopic true color image under the observation of a fluorescence microscope; Step 2: convert, enhance, segment and mark the fluorescence microscopy true color image obtained in step 1: perform grayscale image conversion, adaptive histogram equalization, Otsu's threshold segmentation, and skeleton image feature pixel point marking on the fluorescence microscopy true color image in sequence; Step 3: Based on the Otsu's threshold segmentation image and skeletonized image obtained in step 2, the average swelling diameter ratio, shape factor average value and centroid standard deviation of the polymer modifier are calculated respectively to determine the dispersion uniformity of the polymer modifier in the matrix asphalt.

2. According to claim 1, a specific quantitative method for the dispersion uniformity of a modifier in asphalt based on image processing technology is characterized by: In step 2, the fluorescence microscopy image is successively converted to grayscale, subjected to adaptive histogram equalization, processed by Otsu's threshold segmentation, and skeletonized image feature pixel marking, which are implemented with the assistance of Python code.

3. According to claim 1, a specific quantitative method for the uniformity of dispersion of a modifier in asphalt based on image processing technology is characterized by: In step 3, the average swelling diameter ratio, shape factor average value and centroid standard deviation of the polymer modifier are implemented with the assistance of Python code.

4. The specific quantitative method for the dispersion uniformity of the modifier in asphalt based on image processing technology according to claim 3 is characterized by: According to step 3, The average swelling diameter ratio μD is calculated using the following formula: Among them, T o is the total number of main object pixels, i.e., the total number of polymer modifiers represented by the total number of white pixels; T s is the total number of skeletonized pixels, that is, the marked modifier pixels are processed by the skeletonization algorithm to reduce them to a "skeleton" with a width of one pixel while maintaining the overall structure and connectivity of the modifier.

5. The specific quantitative method for the dispersion uniformity of the modifier in asphalt based on image processing technology according to claim 4 is characterized by: According to step 3, The average shape factor μS is calculated using the following formula: Wherein, A is the area of ​​the white area in the segmented image, that is, the area of ​​the polymer modifier; P is the perimeter of the white area in the segmented image, that is, the perimeter of the polymer modifier.

6. The specific quantitative method for the uniformity of dispersion of a modifier in asphalt based on image processing technology according to claim 5 is characterized by: According to step 3, The centroid standard deviation includes the centroid X-axis standard deviation and the standard deviation of the Y axis of the centroid The following formulas are used for calculation: Among them, x i and i are the x- and y-coordinates of the centroid of the connected region of the ith polymer modifier, and are the average x- and y-coordinates of the centroids of all polymer modifier connected regions, respectively. N is the total number of polymer modifier connected regions, and the centroid is the center position of a modifier connected region.

Images