Bedrock and concrete interface area width detection method based on gray curvature segmentation and fractal driving

Abstract

The invention discloses a bedrock and concrete interface area width detection method based on gray curvature segmentation and fractal driving. The method comprises the following steps: firstly, obtaining a high-resolution two-dimensional profile image of a bedrock-concrete integrated two-medium sample by adopting a computed tomography (CT) technology; a pore gray threshold is crucial to pore recognition, and for the problem that the selection of the current pore gray threshold is relatively large in subjectivity, the pore gray threshold is determined through a gray cumulative distribution curve second derivative method, so that pore structure features are accurately segmented; secondly, constructing a Minkowski dimension model based on a fractal theory, and calculating fractal dimensions of pores of the slice image; and finally, extracting a pore fractal dimension distribution curve along the normal direction of the bedrock-concrete interface, and quantitatively determining the width (Z = Z1-Z0) of an interface transition region according to a fluctuation interval (a starting point Z0 and an ending point Z1) representing interface transition characteristics in the curve.

Description

Technical Field

[0001] The present invention relates to the technical field of the performance of the interface transition zone between bedrock and concrete as a two-medium, and specifically relates to a method for detecting the width of the interface zone between bedrock and concrete based on gray curvature segmentation and fractal driving. Background Art

[0002] As the core load-bearing unit of water conservancy, transportation and other infrastructure, the bedrock-concrete composite structure makes outstanding contributions in aspects such as flood control and power generation, product transportation, and economic development. As the weak part of the bedrock-concrete composite structure, the interface transition zone has prominent problems such as interface cracking and loss of adhesion during long-term service, and even seriously threatens the safety of engineering structures. Therefore, studying the performance of the interface transition zone of the bedrock-concrete composite structure is crucial for ensuring engineering safety. Among them, the quantitative characterization of the width of the interface transition zone is particularly critical for revealing the performance formation mechanism and interface modification optimization. However, there is currently a lack of accurate testing methods.

[0003] Traditional interface transition zone testing technologies mostly focus on the concrete aggregate-mortar system. For example, Chinese invention patents (authorized publication number: CN 108956349B), (publication number: CN 111638103A) use nanoindentation method to indirectly evaluate the width, but there are problems such as irreversible specimen damage and non-repeatability of data; Chinese invention patent (authorized publication number: CN116930237B) realizes non-destructive testing through energy spectrum scanning, but relies on complex pretreatment such as cutting and polishing. In order to identify the width of the interface transition zone between the repair material and concrete, Chinese invention patent (authorized publication number: CN116952995B) uses micro-CT technology to obtain the porosity of each layer of images, and determines the interface transition zone through the change of porosity. However, the determination of its threshold follows the traditional subjective method, and it is difficult to characterize the heterogeneity characteristics of the pore space relying on a single quantitative index, resulting in insufficient spatial resolution of the interface transition zone, thus leading to large measurement errors.

[0004] Based on this, the present invention proposes a method for detecting the width of the interface zone between bedrock and concrete based on gray curvature segmentation and fractal driving. The present invention uses the second derivative method of the gray cumulative curve to avoid the influence of subjective factors, accurately determines the pore gray threshold, and combines the fractal theory to quantify the self-similarity and inhomogeneous distribution characteristics of the pore structure, breaking through the dimensional limitation of the traditional method, and providing an innovative solution with both spatial resolution and morphological sensitivity for the non-destructive and accurate detection of the width of the interface transition zone, solving the problem of discrimination error caused by relying on subjective thresholds in the traditional method and the problem of missing spatial information caused by ignoring the complexity of the pore structure, and improving the accuracy of detecting the width of the interface transition zone. Summary of the Invention

[0005] To solve the problems of poor objectivity in determining the pore gray threshold and failure to consider the complexity of pore structure in the above interface transition zone detection method, the present invention provides a method for detecting the width of the bedrock-concrete interface zone based on gray curvature segmentation and fractal driving. The specific technical solution is as follows:

[0006] Step 1: Pour concrete on the surface of the bedrock, place it in a curing box for curing, and take it out after the bedrock-concrete two-medium sample is cured and formed;

[0007] Step 2: Scan the formed bedrock-concrete two-medium sample in Step 1 with a CT scanner, perform filtering and noise reduction on the obtained 2D cross-sectional image of the sample, and perform enhancement processing, and finally output a high-quality slice image;

[0008] Step 3: Adopt a pore threshold determination algorithm based on the second derivative of the gray cumulative distribution curve, select the first extreme point of the second derivative of the gray cumulative distribution curve as the pore threshold of the slice image, and finally identify and determine the pore phase according to the threshold;

[0009] Step 4: Calculate the fractal dimension of the pores in each slice image using the Minkowski dimension model;

[0010] Step 5: Extract the pore fractal dimension distribution curve along the normal direction of the bedrock-concrete interface. The horizontal axis is the sample height, and the vertical axis is the pore fractal dimension. The interface transition zone between the bedrock and the concrete is defined as the range of drastic fluctuations in the pore fractal dimension distribution curve, and its width Z is determined by the difference between the starting point Z0 and the ending point Z1:

[0011] Z = Z1 - Z0

[0012] In Step 1, the minimum characteristic size of the formed bedrock-concrete two-medium sample must be greater than twice the maximum particle size of the concrete aggregate.

[0013] The method for obtaining the high-quality slice image of the sample in Step 2 is as follows: First, obtain a 2D cross-sectional image with a spatial resolution of no higher than 30 μm / pixel through CT scanning; secondly, use Avizo software to reconstruct the 3D model of the obtained cross-sectional image, cut off the part of the sample affected by edge noise and beam hardening artifacts, and use the Bilateral filtering algorithm to perform noise reduction on the cross-sectional image, and use the Histogram equalization algorithm to eliminate the brightness difference of the cross-sectional image, making it easier to distinguish pores from other phases; finally, output the high-quality slice image of the sample.

[0014] The pore threshold determination algorithm based on the second derivative of the gray cumulative distribution curve in Step 3 is as follows:

[0015]

[0016] F”(g) = F(g + 2) - 2F(g + 1) + F(g)

[0017]

[0018] where g is the gray level in the grayscale image, G max is the maximum gray value, A(k) is the gray area distribution curve, that is, A(k) is the pixel area with gray value k, F(g) is the gray cumulative distribution curve, A total is the total pixel area, F″(g) is the second derivative of the gray cumulative distribution curve, T is the first extreme point of the second derivative of the gray cumulative distribution curve, that is, the pore gray threshold.

[0019] In step four, the Minkowski dimension model is as follows:

[0020] N m,n = Ceil{r -1 [max(g1(m,n), g2(m,n + 1), g3(m + 1,n), g4(m + 1,n + 1))

[0021] - min(g1(m,n), g2(m,n + 1), g3(m + 1,n), g4(m + 1,n + 1)) + 1]}

[0022]

[0023] D = log(N(r)) / log(r -1 )

[0024] where N m,n is the number of boxes required to cover the image in the (m,n) grid, Ceil is the ceiling function, r is the box size, g1, g2, g3, and g4 are the gray values of the four vertices of the grid, N(r) is the total number of boxes required for the entire image, and D is the pore fractal dimension.

[0025] Compared with the prior art, the beneficial effects achieved by the present invention are:

[0026] The present invention proposes a method for detecting the width of the interface area between bedrock and concrete based on gray curvature segmentation and fractal driving. First, a 2D cross-sectional image of a bedrock-concrete integrated two-medium specimen is obtained through CT scanning. The second derivative method of the gray cumulative curve is used to replace the traditional subjective threshold method to objectively determine the pore gray threshold. Secondly, the pore fractal dimension is calculated. By quantifying the self-similarity, inhomogeneous distribution and multi-scale characteristics of the pore structure, the limitations of the traditional porosity index in terms of spatial dimension and morphological sensitivity are broken through. Finally, the pore fractal dimension distribution curve along the interface normal direction is extracted, and the interface transition zone width (Z = Z1 - Z0) is accurately determined based on the statistical fluctuation range (Z0 - Z1) of the curve. This method combines the second derivative method of the gray cumulative curve and fractal theory and applies them to interface detection. While achieving high-precision quantitative characterization of the spatial heterogeneity of the pore structure, it avoids the destruction of the specimen, significantly improves the accuracy of interface recognition and the repeatability of results, effectively solves the recognition error problem caused by relying on subjective thresholds and ignoring the complexity of pores in traditional methods, and provides innovative technical support for the research on the interface performance of hydraulic engineering and geotechnical engineering. Description of the Drawings

[0027] Figure 1 is the flow chart of the present invention;

[0028] Figure 2 is the bedrock surface;

[0029] Figure 3 is the formed bedrock-concrete integrated two-medium specimen;

[0030] Figure 4 is the cross-section of the bedrock-concrete integrated two-medium specimen perpendicular to the interface;

[0031] Figure 5 is the schematic diagram of the steps for obtaining high-quality slice images;

[0032] Figure 6 is the effect diagram for identifying and determining the pore threshold based on the second derivative of the gray cumulative distribution curve;

[0033] Figure 7 is the pore fractal dimension distribution curve extracted along the interface normal direction. Detailed Embodiments

[0034] To make the objectives, technical solutions, and advantages of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below in conjunction with the embodiments and the drawings. It should be noted that the described embodiments are only used to more intuitively explain the core content of the present invention and do not constitute a limitation on the protection scope of the present invention.

[0035] Embodiment 1:

[0036] Method for detecting width of interface zone between bedrock and concrete based on gray curvature segmentation and fractal driving, and specific technical solution is as follows:

[0037] Step 1: Pour concrete on the surface of the bedrock, place it in a curing box for curing, and take it out after the bedrock-concrete integrated two-medium specimen is cured and formed;

[0038] Step 2: Scan the bedrock-concrete integrated two-medium specimen formed in Step 1 through a CT scanner, perform filtering and noise reduction on the obtained 2D sectional image of the specimen, and perform enhancement processing, and finally output a high-quality slice image;

[0039] Step 3: Adopt a pore threshold determination algorithm based on the second derivative of the gray cumulative distribution curve, select the first extreme point of the second derivative of the gray cumulative distribution curve as the pore threshold of the slice image, and finally identify and determine the pore phase according to the threshold;

[0040] Step 4: Calculate the fractal dimension of the pores in each slice image by using the Minkowski dimension model;

[0041] Step 5: Extract the pore fractal dimension distribution curve along the normal direction of the bedrock-concrete interface. The horizontal axis is the specimen height, and the vertical axis is the pore fractal dimension. The interface transition zone between the bedrock and the concrete is defined as the range of drastic fluctuations in the pore fractal dimension distribution curve, and its width Z is determined by the difference between the starting point Z0 and the ending point Z1:

[0042] Z = Z1 - Z0

[0043] The minimum characteristic size of the bedrock-concrete integrated two-medium specimen after forming in Step 1 must be greater than 2 times the maximum particle size of the concrete aggregate.

[0044] The method for obtaining high-quality slice images of the specimen in Step 2 is as follows: First, obtain a 2D sectional image with a spatial resolution not higher than 30 μm / pixel through CT scanning; secondly, use Avizo software to perform 3D model reconstruction on the obtained sectional image, crop the part of the specimen affected by edge noise and beam hardening artifacts, and use the Bilateral filtering algorithm to perform noise reduction processing on the sectional image, and use the Histogram equalization algorithm to eliminate the brightness difference of the sectional image, making it easier to distinguish pores from other phases; finally, output high-quality slice images of the specimen.

[0045] The pore threshold determination algorithm based on the second derivative of the gray cumulative distribution curve in Step 3 is as follows:

[0046]

[0047] F”(g) = F(g + 2) - 2F(g + 1) + F(g)

[0048]

[0049] where g is the gray level in the grayscale image, G max is the maximum gray value, A(k) is the gray area distribution curve, that is, A(k) is the pixel area with gray value k, F(g) is the gray cumulative distribution curve, A total is the total pixel area, F″(g) is the second derivative of the gray cumulative distribution curve, and T is the first extreme point of the second derivative of the gray cumulative distribution curve, that is, the pore gray threshold.

[0050] In step four, the Minkowski dimension model is as follows:

[0051] N m,n = Ceil{r -1 [max(g1(m,n), g2(m,n + 1), g3(m + 1,n), g4(m + 1,n + 1))

[0052] - min(g1(m,n), g2(m,n + 1), g3(m + 1,n), g4(m + 1,n + 1)) + 1]}

[0053]

[0054] D = log(N(r)) / log(r -1 )

[0055] where N m,n is the number of boxes required to cover the image in the (m,n) grid, Ceil is the ceiling function, r is the box size, g1, g2, g3, and g4 are the gray values of the four vertices of the grid, N(r) is the total number of boxes required for the entire image, and D is the pore fractal dimension.

[0056] Example 2:

[0057] This example provides a method for detecting the width of the bedrock - concrete interface area based on gray curvature segmentation and fractal driving. The steps of using this method are described below:

[0058] Step 1: Pour the concrete on the surface of the bedrock, place it in a curing box for curing, and take it out after the bedrock - concrete one - body two - medium specimen is cured and formed. [[ID=4�]]

[0059] Step 2: Scan the bedrock - concrete one - body two - medium specimen formed in step 1 through a CT scanner, perform filtering and noise reduction on the obtained 2D cross - sectional image of the specimen, and perform enhancement processing, and finally output a high - quality sliced image.

[0060] Step 3: Adopt the pore threshold determination algorithm based on the second derivative of the gray cumulative distribution curve, select the first extreme point of the second derivative of the gray cumulative distribution curve as the pore threshold of the slice image, and finally identify the pore phase according to the threshold determination;

[0061] Step 4: Calculate the fractal dimension of the pores in each slice image by using the Minkowski dimension model;

[0062] Step 5: Extract the fractal dimension distribution curve of the pores along the normal direction of the bedrock-concrete interface, where the horizontal axis is the specimen height and the vertical axis is the pore fractal dimension. The interface transition zone between the bedrock and the concrete is defined as the range with severe fluctuations in the fractal dimension distribution curve of the pores, and its width Z is determined by the difference between the starting point Z0 and the ending point Z1: Z = Z1 - Z0.

[0063] Example 3:

[0064] In this example, the method in Example 1 is adopted to detect the width of the interface transition zone between the bedrock and the concrete.

[0065] Pour the concrete on the surface of the bedrock matrix. The bedrock matrix is as Figure 2 shown. Subsequently, place it in a constant temperature and humidity curing box (temperature 20 ± 2°C, relative humidity ≥ 95%) for standard wet curing for 28 days. After the bedrock-concrete is cured and formed, take it out. The bedrock-concrete one-piece two-medium specimen is as Figure 3 shown, and its cross-section perpendicular to the interface is as Figure 4 shown.

[0066] First, use a German CT Precisions microfocus X-ray CT scanner to perform non-destructive tomography scanning on the above-mentioned formed bedrock-concrete one-piece two-medium specimen to obtain a 2D cross-section image of the specimen along the interface normal direction. The spatial resolution of the cross-section image is 30 μm / pixel, and the number of pixels is 1976 × 1976. Secondly, use Avizo software to reconstruct the three-dimensional model of the obtained cross-section image, cut off the part of the specimen affected by edge noise and beam hardening artifacts, and use the Bilateral algorithm to perform noise reduction processing on the cross-section image, and use the Histogram algorithm to eliminate the brightness difference of the cross-section image to make it easier to distinguish the pores from other phases. Finally, output high-quality slice images of the specimen. The schematic diagram of the steps for obtaining high-quality slice images is as Figure 5 shown.

[0067] To avoid the influence of subjective factors and accurately determine the pore gray threshold, the pore gray threshold determination algorithm based on the second derivative of the gray area cumulative distribution curve is adopted. The first extreme point of the second derivative of the gray cumulative distribution curve is selected as the pore gray threshold of the slice image. After calculation, the pore gray threshold of the slice image is 125.71. Finally, the pore phase is identified according to the threshold, as Figure 6 shown.

[0068] The Minkowski dimension model is used to calculate the pore fractal dimension of each slice image. The scale parameter r is set to 30μm - 2000μm, and the number of boxes N(r) containing pores at each r value is counted; a linear regression model is constructed through double logarithmic transformation (logN(r) and log(r -1 )) to characterize the dimension value with the fitting slope. The pore fractal dimension distribution curve along the normal direction of the bedrock-concrete interface is extracted, as Figure 7 shown. The width of the interface transition zone Z = Z1 - Z0 = 3.04mm is determined through the significant fluctuation interval (starting point Z0 = 14.52mm to ending point Z1 = 17.56mm).

Claims

1. A method for detecting the width of the interface zone between bedrock and concrete based on gray curvature segmentation and fractal driving, characterized in that: It includes the following steps: Step 1: Pour concrete onto the surface of the bedrock, place it in a curing box for curing, and take it out after the bedrock-concrete integrated two-medium specimen is cured and formed. Step 2: Scan the formed bedrock-concrete integrated two-medium specimen in Step 1 through a CT scanner, perform filtering and noise reduction on the obtained 2D sectional image of the specimen, and perform enhancement processing, and finally output a high-quality sliced image. Step 3: Adopt a pore threshold determination algorithm based on the second derivative of the gray cumulative distribution curve, select the first extreme point of the second derivative of the gray cumulative distribution curve as the pore threshold of the sliced image, and finally identify and determine the pore phase according to the threshold. Step 4: Calculate the fractal dimension of the pores in each sliced image using the Minkowski dimension model. Step 5: Extract the pore fractal dimension distribution curve along the normal direction of the bedrock-concrete interface, where the horizontal axis is the specimen height and the vertical axis is the pore fractal dimension. The interface transition zone between the bedrock and the concrete is defined as the range of sharp fluctuations in the pore fractal dimension distribution curve, and its width Z is determined by the difference between the starting point Z0 and the ending point Z1: Z = Z1 - Z0.

2. The method for detecting the width of the bedrock and concrete interface area based on gray curvature segmentation and fractal driving according to claim 1, wherein: The minimum characteristic size of the formed bedrock-concrete integrated two-medium specimen in Step 1 must be greater than twice the maximum particle size of the concrete aggregate.

3. The method for detecting the width of the interface zone between bedrock and concrete based on gray curvature segmentation and fractal driving according to claim 1, wherein: The method for obtaining high-quality sliced images of the specimen in Step 2 is as follows: First, obtain a 2D sectional image with a spatial resolution of no higher than 30 μm / pixel through CT scanning; secondly, use Avizo software to reconstruct the three-dimensional model of the obtained sectional image, cut off the part of the specimen affected by edge noise and beam hardening artifacts, and use the bilateral filtering algorithm to perform noise reduction processing on the sectional image, and use the histogram equalization algorithm to eliminate the brightness difference of the sectional image, making it easier to distinguish pores from other phases; finally, output high-quality sliced images of the specimen.

4. The method for detecting the width of the bedrock-concrete interface area based on gray curvature segmentation and fractal driving according to claim 1, wherein: The pore threshold determination algorithm based on the second derivative of the gray cumulative distribution curve in Step 3 is as follows: g ∈ {0, 1, ..., G max} F”(g) = F(g + 2) - 2F(g + 1) + F(g) where g is the gray level in the grayscale image, G max is the maximum gray value, A(k) is the gray area distribution curve, that is, A(k) is the pixel area with the gray value of k, F(g) is the gray cumulative distribution curve, A total is the total pixel area, F″(g) is the second derivative of the gray cumulative distribution curve, T is the first extreme point of the second derivative of the gray cumulative distribution curve, that is, the pore gray threshold.

5. The method for detecting the width of the bedrock-concrete interface zone based on gray curvature segmentation and fractal driving according to claim 1, wherein: The Minkowski dimension model in Step 4 is as follows: N m,n = Ceil{r -1 [max(g1(m,n), g2(m,n + 1), g3(m + 1,n), g4(m + 1,n + 1)) -min(g1(m,n), g2(m,n + 1), g3(m + 1,n), g4(m + 1,n + 1)) + 1]} D = log(N(r)) / log(r -1 ) where N m,n is the number of boxes required to cover the image in the (m,n) grid, Ceil is the ceiling function, r is the box size, g1, g2, g3, and g4 are the gray values of the four vertices of the grid, N(r) is the total number of boxes required for the entire image, and D is the pore fractal dimension.

Images