A digital image processing method for asphalt mixture based on partition processing

Abstract

This invention discloses a method for processing digital images of asphalt mixtures based on partitioning, comprising the following steps: 1. obtaining a raw CT grayscale image of the asphalt mixture; 2. converting the raw CT image into a grayscale image with a grayscale range of 0 to 255, and determining the center and radius of a circle in a two-dimensional cross-sectional image of the asphalt mixture specimen; 3. dividing the digital image into a circle and several rings with uniform grayscale according to its grayscale distribution; 4. performing multi-threshold Otsu segmentation on the resulting circles and rings to obtain binary images of the aggregate; and 5. sequentially assembling the binary images of the circles and rings obtained through the multi-threshold Otsu segmentation to obtain a digital image of the asphalt mixture containing complete aggregate information. This method eliminates the effects of uneven grayscale distribution and ring artifacts on the extraction of internal microstructural information of the asphalt mixture.

Description

Technical Field

[0001] The present invention belongs to the technical field of road engineering, and in particular relates to an asphalt mixture digital image processing method based on partition processing. Background Art

[0002] Asphalt mixtures are multiphase composite materials composed of aggregate, filler, and asphalt binder, with a certain amount of voids. Their strength evolution is inherently correlated with the dynamic characteristics of their skeleton structure. However, existing research has mostly evaluated the structural strength of asphalt mixtures based on macroscopic volumetric metrics, which cannot fully explain the mechanisms underlying their strength development. Therefore, it is necessary to investigate the strength development mechanisms of asphalt mixtures from a microscopic perspective to provide more precise guidance for asphalt pavement construction and quality evaluation.

[0003] With the development of computer and image processing technology, patents such as Patent 202210126323X have begun to use X-ray CT scanning to obtain internal structural information of asphalt mixtures. The basic step is to process the two-dimensional digital image of the asphalt mixture and extract characteristic information such as aggregates and voids to study the strength formation law of the asphalt mixture. However, the binary image of the aggregate in the ring in the above patent will have varying degrees of adhesion. Mortar with similar grayscale values ​​to aggregates will be mistaken for aggregates, which will cause large areas of adhesion in the final two-dimensional aggregate digital image. The reason for this phenomenon is that the ring division cannot eliminate the impact of "artifacts" caused by X-ray intensity attenuation on the asphalt mixture CT image, which in turn causes the aggregate segmentation threshold to be too large or too small, ultimately resulting in large areas of adhesion in the aggregate binary image. Summary of the Invention

[0004] The main purpose of the present invention is to provide an asphalt mixture digital image processing method based on partitioning processing, an electronic device and a computer-readable storage medium, aiming to eliminate the influence of uneven grayscale distribution of asphalt mixture CT digital images on the segmentation threshold and realize accurate segmentation of aggregate particles.

[0005] To this end, an embodiment of the present invention provides a method for processing a digital image of an asphalt mixture based on partition processing, comprising the following steps:

[0006] Step 1: Obtaining a CT grayscale image of the original asphalt mixture;

[0007] Step 2: convert the original CT image into a grayscale image with a grayscale range of 0 to 255, and determine the center and radius of the circle in the two-dimensional cross-sectional image of the asphalt mixture specimen;

[0008] Step 3: Divide the digital image into a circle and several rings with uniform grayscale according to its grayscale distribution;

[0009] Step 4: Based on the circle and ring images divided in step 3, the proportion of the number of pixels in the foreground and background of each circle or ring is counted;

[0010] Based on the obtained ratio of the total number of pixels of the foreground part and the background part, the average grayscale size of the foreground part and the background part is calculated;

[0011] Based on the obtained ratio of the number of pixels in the background part, the ratio of the number of pixels in the foreground part, the average gray value of the foreground part, and the average gray value of the background part, the multi-threshold Otsu segmentation method is adopted. With 0 as the starting threshold, the threshold T1 is calculated according to the maximum inter-class variance. Then, with T1 as the starting threshold, the above steps are repeated to calculate the threshold T2. With T2 as the segmentation threshold, the image is segmented to obtain a circular image containing only aggregate information.

[0012] Step 5: Assemble the binary circle and ring images obtained by multi-threshold Otsu segmentation in sequence to obtain a digital image of the asphalt mixture containing complete aggregate information.

[0013] Specifically, in step 2, the original CT image is converted into a grayscale image with a grayscale range of 0 to 255, and the center and radius of the circle in the two-dimensional cross-sectional image of the asphalt mixture specimen are determined; the specific process is as follows:

[0014] In the asphalt mixture digital image: the cross-section image of the target specimen is circular;

[0015] Step 1: Convert the asphalt mixture digital image into an 8-bit grayscale image, that is, a grayscale image with a grayscale range of 0 to 255, which can improve the computer processing speed;

[0016] Step 2: Use the bwboundaries function in Matlab software to track the boundary area of ​​the target image and determine the center and radius of the minimum inscribed circle of the target area of ​​the image;

[0017] Specifically, in step 3, the target area with the determined center and radius is divided into a circle and several rings with uniform grayscale distribution according to the uniformity of grayscale distribution. The specific process is as follows:

[0018] According to the center and radius of the minimum inscribed circle determined in step 2, the target area is divided into a circle and an annulus. The mathematical expression of the annulus in the image is shown in (1):

[0019]

[0020] Where a0 and b0 are the horizontal and vertical coordinates of the center of the minimum inscribed circle of the target area, a and b are the horizontal and vertical coordinates of the pixel points on the ring, and k i is the outer diameter of the i-th ring, ki-1 is the outer diameter of the i-1th ring. When i is 1, the above formula is the innermost circle of the target area.

[0021] Specifically, based on the circle and ring images divided in step 3, the ratio of the number of background pixels to the number of foreground pixels is calculated with each circle or ring as an independent unit. The calculation formula is as follows:

[0022] P i1 =N i1 / Sall i (2)

[0023] P i2 =N i2 / Sall i (3)

[0024] Where N i1 is the number of background pixels on the i-th ring, N i2 is the number of foreground pixels on the i-th ring, Sall i is the total number of pixels on the i-th ring, P i1 is the ratio of the total number of background pixels to the total number of ring pixels in the i-th ring, P i2 is the ratio of the total number of foreground pixels to the total number of pixels in the ring in the i-th ring;

[0025] After obtaining the proportion of the total number of pixels in the foreground and background parts, the average grayscale size of the foreground and background parts is calculated. The calculation formula is as follows:

[0026] Ave i1 =M i1 / N i1 (4)

[0027] Ave i2 =M i2 / N i2 (5)

[0028] Where M i1 is the sum of the grayscale values ​​of the background pixels on the i-th ring, M i2 is the sum of the grayscale values ​​of the pixels belonging to the foreground on the i-th ring, Ave i1 is the average gray value of the background part of the i-th ring, Ave i2 is the average gray value of the foreground part of the i-th ring.

[0029] Specifically, the circular images containing aggregate information obtained in step 4 are assembled to obtain a complete digital image of the asphalt mixture, wherein the assembly formula is as follows:

[0030] f(x,y)=f1(x,y)+f2(x,y)+…………+f i (x, y) (6)

[0031] Where f1(x, y) is the grayscale distribution function of the first circle containing only aggregate information, f2(x, y) is the grayscale distribution function of the second circle containing only aggregate information, and f i (x, y) is the grayscale distribution function of the i-th circle containing only aggregate information, and f(x, y) is the grayscale distribution function of the entire digital image containing only aggregate information.

[0032] Specifically, the digital image containing aggregate information obtained in step five is subjected to median filtering and morphological processing to remove noise and fine particles in the image.

[0033] Another embodiment of the present invention provides an electronic device including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor implements the above-mentioned asphalt mixture digital image processing method.

[0034] Another aspect of an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the computer program implements the above-mentioned asphalt mixture digital image processing method.

[0035] Compared with the prior art, the present invention has the following beneficial effects: in the circular ring image obtained by processing using the present invention, there is basically no adhesion between the aggregate, mortar and voids, the aggregate boundary is clear, and the contour is well preserved, eliminating the influence of problems such as uneven grayscale distribution and ring artifacts on the extraction of internal micro-structural information of asphalt mixture, solving the problem of poor adaptability of traditional digital image processing methods, and being suitable for large-scale processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0037] Figure 1 This is a flowchart of image processing of the present invention;

[0038] Figure 2(a) shows a CT digital image of asphalt mixture with a grayscale range of 0 to 255 after format conversion. Figure 2(b) shows a schematic diagram of automatically determining the image center coordinates and radius using the bwboundaries function in Matlab software.

[0039] Figure 3 Schematic diagram of the annular partitioning of asphalt mixture CT images;

[0040] 4(a) to (f) are, from left to right, the original grayscale image after annular segmentation, the annular single threshold segmentation effect diagram, and the partition processing effect diagram of the double OTSU segmentation of the present invention;

[0041] FIG5( a ) is a diagram showing the effect of the ring combination after the present invention has been processed, and FIG5( b ) is a diagram showing the effect of FIG5( a ) after morphological processing;

[0042] Figure 6 This is the combined effect diagram of the ring single threshold segmentation;

[0043] Figure 7 This is the effect diagram based on global threshold segmentation. DETAILED DESCRIPTION

[0044] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0045] See also Figure 1 A method for processing digital images of asphalt mixtures based on partition processing comprises the following steps:

[0046] Step 1: Obtain the original asphalt mixture CT grayscale image

[0047] Step 2: Convert the original CT image into a grayscale image with a grayscale range of 0 to 255, and determine the center and radius of the circle in the two-dimensional cross-sectional image of the asphalt mixture specimen, as shown in Figure 2(a);

[0048] In the asphalt mixture digital image: the cross-section image of the target specimen is circular;

[0049] Step 1: Convert the asphalt mixture digital image into an 8-bit grayscale image, that is, a grayscale image with a grayscale range of 0 to 255, which can improve the computer processing speed;

[0050] Step 2: Use the bwboundaries function in Matlab software to track the boundary area of ​​the target image and determine the center and radius of the minimum inscribed circle of the target area of ​​the image, as shown in Figure 2(b).

[0051] Step 3: Divide the digital image into a circle and several rings with uniform grayscale according to its grayscale distribution

[0052] According to the center and radius of the minimum inscribed circle determined in step 2, the target area is divided into circles and rings, such as Figure 3 As shown, the mathematical expression of the ring in the image is shown in (1):

[0053]

[0054] Where a0 and b0 are the horizontal and vertical coordinates of the center of the minimum inscribed circle of the target area, a and b are the horizontal and vertical coordinates of the pixel points on the ring, and k i is the outer diameter of the i-th ring, k i-1 is the outer diameter of the i-1th ring. When i is 1, the above formula is the innermost circle of the target area.

[0055] Step 4: Based on the circle and ring images divided in step 3, count the proportion of the number of pixels in the foreground and background of each circle or ring as an independent unit. The calculation formula is as follows:

[0056] P i1 =N i1 / Sall i (2)

[0057] P i2 =N i2 / Sall i (3)

[0058] Where N i1 is the number of background pixels on the i-th ring, N i2 is the number of foreground pixels on the i-th ring, Sall i is the total number of pixels on the i-th ring, P i1 is the ratio of the total number of background pixels to the total number of ring pixels in the i-th ring, P i2 is the ratio of the total number of foreground pixels to the total number of ring pixels in the i-th ring;

[0059] Based on the obtained ratio of the total number of pixels of the foreground and background parts, the average grayscale size of the foreground and background parts is calculated. The calculation formula is as follows:

[0060] Ave i1 =M i1 / N i1 (4)

[0061] Ave i2 =M i2 / N i2 (5)

[0062] Where M i1 is the sum of the grayscale values ​​of the background pixels on the i-th ring, Mi2 is the sum of the grayscale values ​​of the pixels belonging to the foreground on the i-th ring, Ave i1 is the average gray value of the background part of the i-th ring, Ave i2 is the average gray value of the foreground part of the i-th ring;

[0063] Based on the obtained ratio of the number of pixels in the background part, the ratio of the number of pixels in the foreground part, the average gray value of the foreground part and the average gray value of the background part, the multi-threshold Otsu segmentation method is adopted. With 0 as the starting threshold, the threshold T1 is calculated according to the maximum inter-class variance. Then, with T1 as the starting threshold, the above steps are repeated to calculate the threshold T2. With T2 as the segmentation threshold, the image is segmented to obtain a circular image containing only aggregate information, as shown in Figures 4(a) to (f); where,

[0064] Specifically, the principle of the OTSU segmentation method is to maximize the maximum inter-class variance of the target part (foreground part) and the background part of the image, as shown in formula (1):

[0065]

[0066] w0 is the ratio of the number of background pixels;

[0067] w1 is the proportion of the number of pixels in the target part;

[0068] μ0 is the average gray value of the background part;

[0069] μ1 is the average gray value of the target part.

[0070] For an 8-bit asphalt mixture grayscale image, the specific steps of OTSU threshold segmentation are as follows: the value of T ranges from 0 to 255, and the inter-class variance obtained in each cycle is calculated according to formula (1): The T corresponding to the maximum inter-class variance is taken as the image segmentation threshold to obtain the binary image of the aggregate.

[0071] Step 5: The binary circle and ring images obtained by multi-threshold Otsu segmentation are assembled in sequence to obtain a digital image of the asphalt mixture containing complete aggregate information, as shown in Figure 5(a). The obtained digital image containing aggregate information is then subjected to median filtering and morphological processing to obtain an image with noise and fine particles removed, as shown in Figure 5(b).

[0072] The assembly formula is as follows:

[0073] f(x,y)=f1(x,y)+f2(x,y)+···········+f i (x, y) (6)

[0074] Where f1(x, y) is the grayscale distribution function of the first circle containing only aggregate information, f2(x, y) is the grayscale distribution function of the second circle containing only aggregate information, and f i (x, y) is the grayscale distribution function of the i-th circle containing only aggregate information, and f(x, y) is the grayscale distribution function of the entire digital image containing only aggregate information.

[0075] Figure 6 The figure shows the effect of an OTSU segmentation obtained using the method described in patent 202210126323X. Figure 7 Figure 5(b) shows the ring segmentation renderings obtained using the present invention. By comparison, the ring images obtained in this application show essentially no adhesion between the aggregate, mortar, or voids. The final renderings demonstrate clear aggregate boundaries and well-preserved contours. Overall, the automatic multi-threshold ring segmentation proposed in this invention eliminates the impact of uneven grayscale distribution in asphalt mixture CT images on the segmentation threshold, achieving accurate segmentation of aggregate particles.

[0076] Another embodiment of the present invention provides an electronic device including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor implements the above-mentioned asphalt mixture digital image processing method.

[0077] Another aspect of an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the computer program implements the above-mentioned asphalt mixture digital image processing method.

[0078] The above embodiments are merely examples to clearly illustrate the present invention and are not intended to limit its implementation. Those skilled in the art will readily appreciate that other variations or modifications based on the above descriptions are possible. It is not necessary and impossible to provide an exhaustive list of all embodiments. Obvious variations or modifications arising therefrom remain within the scope of protection of the present invention.

Claims

1. A method for digital image processing of asphalt mixture based on partition processing, characterized in that: All steps are completed by computer operation. The specific process of the method is as follows: Step 1: Obtaining a CT grayscale image of the original asphalt mixture; Step 2: convert the original CT image into a grayscale image with a grayscale range of 0 to 255, and determine the center and radius of the circle in the two-dimensional cross-sectional image of the asphalt mixture specimen; Step 3: Divide the digital image into a circle and several rings with uniform grayscale according to its grayscale distribution; Step 4: Based on the circle and ring images divided in step 3, the proportion of the number of pixels in the foreground and background of each circle or ring is counted; Based on the obtained ratio of the total number of pixels of the foreground part and the background part, the average grayscale size of the foreground part and the background part is calculated; Based on the obtained ratio of the number of pixels in the background part, the ratio of the number of pixels in the foreground part, the average gray value of the foreground part, and the average gray value of the background part, the multi-threshold Otsu segmentation method is adopted. With 0 as the starting threshold, the threshold T1 is calculated according to the maximum inter-class variance. Then, with T1 as the starting threshold, the above steps are repeated to calculate the threshold T2. With T2 as the segmentation threshold, the image is segmented to obtain a circular image containing only aggregate information. Step 5: Assemble the binary circle and ring images obtained by multi-threshold Otsu segmentation in sequence to obtain a digital image of the asphalt mixture containing complete aggregate information.

2. The asphalt mixture digital image processing method based on partition processing according to claim 1 is characterized in that: The specific process of step 2 is: Step 1: Convert the asphalt mixture digital image into an 8-bit grayscale image; Step 2: Use the bwboundaries function in Matlab software to track the boundary area of ​​the target image and determine the center and radius of the minimum inscribed circle of the target area of ​​the image.

3. The asphalt mixture digital image processing method based on partition processing according to claim 1 is characterized in that: The specific process of step three is: According to the center and radius of the minimum inscribed circle determined in step 2, the target area is divided into a circle and an annulus. The mathematical expression of the annulus in the image is shown in (1): Where a0 and b0 are the horizontal and vertical coordinates of the center of the minimum inscribed circle of the target area, a and b are the horizontal and vertical coordinates of the pixel points on the ring, and k i is the outer diameter of the i-th ring, k i-1 is the outer diameter of the i-1th ring. When i is 1, the above formula is the innermost circle of the target area.

4. The asphalt mixture digital image processing method based on partition processing according to claim 1 is characterized in that: The formula for calculating the ratio of the total number of pixels in the foreground and background is as follows: P i1 =N i1 / Sall i (2) P i2 =N i2 / Sall i (3) Where N i1 is the number of background pixels on the i-th ring, N i2 is the number of foreground pixels on the i-th ring, Sall i is the total number of pixels on the i-th ring, P i1 is the ratio of the total number of background pixels to the total number of ring pixels in the i-th ring, P i2 is the ratio of the total number of foreground pixels to the total number of ring pixels in the i-th ring.

5. The asphalt mixture digital image processing method based on zoning processing according to claim 1 is characterized in that: The calculation formula for the average grayscale size of the foreground and background parts is as follows: Ave i1 =M i1 / N i1 (4) Ave i2 =M i2 / N i2 (5) Where M i1 is the sum of the grayscale values ​​of the background pixels on the i-th ring, M i2 is the sum of the grayscale values ​​of the pixels belonging to the foreground on the i-th ring, Ave i1 is the average gray value of the background part of the i-th ring, Ave i2 is the average gray value of the foreground part of the i-th ring.

6. The asphalt mixture digital image processing method based on partition processing according to claim 1 is characterized in that: The circular images containing aggregate information obtained in step 4 are assembled to obtain a complete digital image of the asphalt mixture. The assembly formula is as follows: f(x,y)=f1(x,y)+f2(x,y)+···········+f i (x,y) (6) Where f1(x, y) is the grayscale distribution function of the first circle containing only aggregate information, f2(x, y) is the grayscale distribution function of the second circle containing only aggregate information, and f i (x, y) is the grayscale distribution function of the i-th circle containing only aggregate information, and f(x, y) is the grayscale distribution function of the entire digital image containing only aggregate information.

7. The asphalt mixture digital image processing method based on partition processing according to claim 1 is characterized in that: The digital image containing aggregate information obtained in step five is subjected to median filtering and morphological processing to remove noise and fine particles in the image.

8. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the computer program is executed by the processor, the processor implements the asphalt mixture digital image processing method according to any one of claims 1 to 7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the asphalt mixture digital image processing method according to any one of claims 1 to 7 is implemented.

Images