A core fracture parameter detection method based on multi-scale local contrast measure

Abstract

The present application belongs to the technical field of oil and gas exploration and development and image processing, and relates to a core fracture parameter detection method. After the original image of the core fracture is obtained, the scale in the image is read, and the conversion relationship between the pixel size and the actual size is established; the original image is corrected for illumination to obtain an enhanced image; the internal region and the external region are constructed in the enhanced image, the local contrast under multiple scales is calculated and fused to obtain a fused contrast graph; the dark channel graph is calculated according to the dark channel prior, and is combined with the fused contrast graph to synthesize a fracture saliency graph; the fracture saliency graph is subjected to OTSU threshold segmentation, closed operation and connected region screening to obtain a fracture binary graph; the fracture region is subjected to skeletonization processing, the fracture length is calculated according to the number of skeleton pixels and the proportionality coefficient, and the average opening degree and the surface density are calculated in combination with the fracture area. The method can be used for fracture region extraction in a core photo and fracture parameter statistics.

Description

Technical Field

[0001] This invention relates to the fields of oil and gas exploration and development and image processing technology, and in particular to an image detection method for identifying fracture regions in rock cores and calculating fracture length, average aperture, and areal density. Background Technology

[0002] In the field of oil and gas exploration and development, the degree of reservoir fracture development directly affects oil and gas migration efficiency and ultimate recovery rate. Accurate acquisition of fracture parameters is the foundation for reservoir evaluation and optimization of development plans. Currently, fracture detection and parameter extraction mainly rely on four approaches: production dynamic analysis indirectly infers the existence of fractures from production data, but cannot provide quantitative parameters; geostress analysis predicts fracture distribution based on mechanical mechanisms, but is constrained by theoretical assumptions and experimental conditions, resulting in significant estimation errors; imaging logging can obtain downhole fracture information, but the measurement cost is high and the cycle is long, making it difficult to carry out large-scale operations in non-key well locations; core fracture observation directly observes the core of the target layer to obtain fracture parameters, and its results are the most accurate and reliable, making it one of the most widely used methods in the field. However, core fracture observation faces two prominent problems in practical operation: first, the fractures and background overlap and become messy in the original fracture images, requiring manual identification based on experience; second, the manual measurement and statistics of fracture parameters involve a large amount of repetitive work, and the manual processing efficiency is low when processing hundreds or thousands of core images. Therefore, for batches of core images, there is still a need for a method to extract crack parameters that can reduce the workload of manual line drawing and manual statistical analysis.

[0003] To address the aforementioned issues, existing research has attempted to incorporate digital image processing techniques into core fracture detection. For example, edge detection operators are used to extract fracture contours, or threshold segmentation methods are employed to separate fracture regions from the background. However, actual core images reveal complex fracture morphologies and significant width variations. Furthermore, rock surfaces often contain various non-crack structures such as mineral spots, scratches, and gravel, whose color and texture characteristics overlap with fractures to some extent. This leads to fractures being easily broken or falsely detected, affecting subsequent calculations of parameters such as length, aperture, and areal density. Therefore, in core fracture image processing, how to stably extract fracture regions under complex rock textures, uneven illumination, and local noise interference, and further calculate fracture length, aperture, and areal density, remains a problem that needs to be solved. Summary of the Invention

[0004] To address the problems of discontinuous fracture boundaries, significant background texture interference, and time-consuming manual parameter calculation in core images, this invention provides a method for detecting fracture parameters in core images. This method includes the following steps: Acquire the original image of the rock core fracture, identify the scale bar in the image, and calculate the scaling factor between the pixel size and the actual size; The original image is subjected to illumination correction and enhancement processing to obtain an enhanced image with relatively uniform brightness distribution; Centered on the pixel in the enhanced image, construct an inner region R1 and an outer region R2 respectively. Calculate the difference in the feature mean of the two regions in the CIE Lab color space to obtain the local contrast value of the pixel at the corresponding scale. Local contrast calculations are repeated at multiple external region scales, and the detection maps at different scales are fused to obtain a fused contrast map. The dark channel prior map is calculated based on the enhanced image, and the dark channel prior map is weighted and fused with the fused contrast map to obtain the crack saliency map; An adaptive threshold segmentation is performed on the crack saliency map to obtain an initial binary crack map; Morphological closing operations are performed on the initial binary crack map, and scattered or blocky non-crack regions are removed according to the ratio of the major axis to the minor axis of the connected region to obtain the final crack region. The final crack area is ossified, the crack length is calculated using a proportional coefficient, and the average crack aperture and crack surface density are further calculated based on the crack area.

[0005] The image enhancement process can employ an enhancement method based on a light reflection model and a multi-scale Gaussian function; the threshold segmentation can use OTSU adaptive threshold segmentation; and the crack length can be calculated from the number of pixels in the ossified crack skeleton.

[0006] Compared with existing technologies, this invention provides a method for detecting core fracture parameters based on multi-scale local contrast measurement, which has the following advantages: 1. Before crack identification, this invention establishes the correspondence between pixel size and actual object size. Subsequently, crack length, average aperture and surface density can be converted according to the same proportional coefficient, avoiding manual repeated conversion and item-by-item statistics.

[0007] 2. This invention uses multiple neighborhood scales to calculate local contrast. Small-scale results are used to suppress local background interference, while large-scale results are used to preserve crack continuity. After fusion, the impact of crack discontinuity or false background detection at a single scale can be reduced.

[0008] 3. This invention fuses the dark channel prior image with the local contrast image, so that the differences in crack edges and the features of dark areas of cracks can be judged at the same time. This is beneficial to preserve the main area of ​​the crack and reduce crack breakage caused by relying solely on edge information.

[0009] 4. After binarization, this invention uses closing operation and the ratio of major and minor axes of connected regions for screening, which can fill in some discontinuous areas of cracks and remove some non-crack areas that are approximately blocky or scattered, providing a binary map basis for subsequent ossification processing and crack area calculation. Attached Figure Description

[0010] Figure 1 The original core reservoir fracture image provided in the embodiments of the present invention; Figure 2 This is an overall flowchart of the core fracture parameter detection method provided in the embodiments of the present invention; Figure 3 This is an image of a core reservoir fracture after image enhancement processing, provided in an embodiment of the present invention. Figure 4 This is a schematic diagram of the inner and outer regions of a local contrast filter provided in an embodiment of the present invention, wherein... Figure 4 (a) is a schematic diagram of a local contrast filter composed of the inner region R1 and the outer region R2. Figure 4 (b) is a schematic diagram when the internal region R1 remains unchanged and the scale of the external region R2 is halved; Figure 5 This is a multi-scale local contrast detection result image provided in an embodiment of the present invention, wherein, Figure 5 (a) shows the local contrast detection results at the first scale. Figure 5 (b) shows the local contrast detection results at the second scale. Figure 5 (c) shows the local contrast detection results at the third scale. Figure 5 (d) is the result of linear weighted fusion of local contrast detection maps at three scales; Figure 6 This is a diagram showing the detection results of core reservoir fractures based on dark channel priors provided in an embodiment of the present invention. Figure 7 The binary image of core reservoir fractures obtained after morphological closure operation and major axis-to-minor axis ratio screening is provided in an embodiment of the present invention. Figure 8 This is a crack skeleton diagram obtained by performing morphological ossification processing on a binary crack image, as provided in an embodiment of the present invention. Detailed Implementation

[0011] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0012] Example 1

[0013] Please see Figure 1-8 This embodiment uses six reservoir core fracture photographs actually collected from an oil and gas field (e.g., Figure 1The images shown are the processing objects, with a resolution of 1200×1600 pixels and a JPEG format. The method of this invention is used to sequentially detect fracture regions and calculate fracture parameters in the aforementioned core fracture images. The specific implementation steps are as follows.

[0014] Step 1: Image Scale Bar Extraction

[0015] After acquiring the original core fracture image, the number of pixels occupied by the reference scale in the image is first extracted. In this embodiment, the core photographs all contain a real-world reference scale of known length. The scale region is identified and its pixel length is calculated using image processing. The ratio of the real-world size to the image pixel size is calculated to be 1:60, meaning one pixel in the image corresponds to 0.0167 cm of the real object. This ratio is stored and used for subsequent conversion of fracture parameters to their actual dimensions.

[0016] Step 2: Image Enhancement Processing

[0017] Due to limitations in on-site shooting conditions, raw core fracture images often suffer from uneven illumination, with some fracture areas exhibiting low brightness, which affects the accuracy of subsequent fracture detection. Therefore, this embodiment employs a nonlinear function transformation method based on an illumination reflection model and multi-scale theory to enhance the raw images. The specific process is as follows.

[0018] (1) Convert the original image from RGB color space to HSV color space and extract the luminance channel component. .

[0019] (2) Based on the illumination reflection model, the illumination components of the brightness channel are estimated using a multi-scale Gaussian function. :

[0020] in, This represents the convolution operation. For the first The expression for the Gaussian function with multiple scales is:

[0021] In this embodiment, three scales (N=3) are used, with Gaussian kernel scale parameters. The weighting coefficients for each scale are 15, 80, and 200 respectively. All are taken as 1 / 3. The normalization factor ensures that the Gaussian function satisfies... .

[0022] (3) Use weighting coefficients respectively and The luminance channel is enhanced to obtain two enhanced luminance images. and The enhancement formula is:

[0023] in, It is a small constant (0.001 in this embodiment) to prevent the denominator from being zero.

[0024] (4) Calculate the covariance matrix of the two enhanced images, solve for their eigenvalues ​​and eigenvectors, and adaptively fuse the two enhanced images according to the coefficients of the eigenvectors to obtain the final enhanced brightness channel.

[0025] (5) The enhanced luminance channel is merged with the saturation and chroma channels in the original HSV space and converted back to the RGB color space to obtain an enhanced image with a more balanced luminance distribution.

[0026] The enhancement results of the six core fracture images in this embodiment are as follows: Figure 3 As shown. By Figure 3 As can be seen, the dark areas in the enhanced image are improved to a certain extent, and the grayscale difference between the crack and the surrounding rock background is more easily utilized in subsequent local contrast calculations.

[0027] Step 3: Crack Detection Based on Multi-Scale Local Contrast

[0028] In the enhanced core fracture image, the fracture area and the surrounding rock background usually differ in color or grayscale distribution. In this embodiment, a multi-scale local contrast method is used to extract the fracture area.

[0029] (1) For each pixel location in the enhanced image, construct an inner region R1 (3×3 pixels) centered on that pixel, and an outer region R2 surrounding R1. The width of region R2 is... Values ​​should be taken from the following range:

[0030] in, These are the scale parameters used to determine the width of the outer region R2. and These represent the height and width of the image, respectively.

[0031] (2) Convert the image from RGB space to CIE Lab color space and extract the L, a, and b channel components. For each pixel location, calculate the mean vector of the inner region R1 and the outer region R2 in the L, a, and b channels respectively. and Then, the Euclidean distance between the two is calculated as the local contrast value at that location:

[0032] (3) In this embodiment, three scales are selected for local contrast calculation, corresponding to Take 5, 9 and 15 respectively (i.e. =2, 4, 7). At each scale, traverse all pixels of the entire image to generate a local contrast detection map at that scale.

[0033] (4) The local contrast detection maps at the three scales are sorted by equal weight. Linear weighted fusion is performed to obtain a fused contrast map. The fusion formula is:

[0034] in For the first Local contrast values ​​at each scale.

[0035] In this embodiment, the local contrast detection results at three scales are as follows: Figure 5 As shown in (a)-(c), the fusion results are as follows: Figure 5 As shown in (d). From Figure 5 It can be seen that the crack display effect varies at different scales: at smaller scales, there is less background response, but some crack segments are not continuous enough; at larger scales, the crack continuity is better, but background areas may also be included. By fusing the results from multiple scales, a compromise can be achieved between crack continuity and background suppression.

[0036] Step 4: Crack Detection Based on Dark Channel Prior

[0037] This embodiment also employs the dark channel prior method to extract the crack region, complementing the local contrast method.

[0038] For enhanced core fracture images Its secret passage Calculate according to the following formula:

[0039] Let c be the components of the input image in color channel c (red, green, blue). This refers to a local image patch centered at pixel x. In this embodiment, the size of the local image patch is 15×15 pixels.

[0040] The theoretical basis for the dark channel prior is that in outdoor, fog-free images, except for the sky area, at least one color channel has some pixel intensity values ​​that are very low and close to zero. In the core image of this embodiment, the crack area usually appears as a darker band, and its dark channel response differs from the surrounding rock background. Therefore, the dark channel map can serve as auxiliary information for crack area extraction.

[0041] The dark channel prior detection results of the six core fracture images in this embodiment are as follows: Figure 6 As shown.

[0042] Step 5: Crack saliency map generation and binarization segmentation

[0043] (1) The fusion contrast map obtained in step three Compared with the dark channel prior map obtained in step four Perform linear weighted fusion to generate a crack saliency map. :

[0044] In the formula, and These are the preset weight coefficients for the fused contrast image Dfusion and the dark channel prior image Jdark, respectively. + =1. The fused contrast map is used to characterize the local differences between the crack region and the surrounding background, while the dark channel prior map is used to supplement the response of the dark crack region. To ensure that local contrast information dominates the fusion result, while preserving the supplementary role of dark channel information for dark cracks, the following settings are configured: Greater than This embodiment follows the above-described setting principles, and before image processing, [the image is processed]. and The values ​​are preset to 0.7 and 0.3 respectively, and remain unchanged during the generation of the crack saliency map.

[0045] (2) The OTSU adaptive threshold segmentation algorithm is used to analyze the crack saliency map. Binarization is performed. The OTSU algorithm automatically determines the optimal segmentation threshold by maximizing the inter-class variance. The image pixels are divided into crack targets (grayscale values ​​greater than 100%). ) and background (grayscale value less than or equal to Two types of cracks are generated to produce an initial binary crack map.

[0046] Step Six: Crack Region Identification and Refinement Based on Mathematical Morphology

[0047] The initial binary crack map contains some non-crack regions caused by noise or image texture (mainly appearing as scattered or small connected regions), which need further processing to reduce the impact of non-crack responses on subsequent parameter calculations.

[0048] (1) Morphological closing operation is used to process the initial binary crack image. The closing operation is an operation of first expanding and then eroding, used to fill small holes in the crack area and connect some discontinuous crack segments. In this embodiment, the structural element is a circular structural element with a radius of 3 pixels.

[0049] (2) Mark the connected regions of the binary graph after the closing operation and obtain all independent connected regions.

[0050] (3) Calculate the length of the major axis of the minimum circumscribed ellipse of each connected region. and minor axis length And calculate the ratio of the two. Cracked areas are typically elongated and narrow, with a major-to-minor axis ratio much greater than 1; while noise points and non-cracked areas (such as mineral spots and scratches on the core surface) are usually nearly circular or square, with a major-to-minor axis ratio close to 1. In this embodiment, Connected regions are identified as non-crack regions and removed, while remaining... The area is designated as the final crack area.

[0051] The final binary fracture maps of the six core fracture images in this embodiment are as follows: Figure 7 As shown. From Figure 7 It can be seen that after the closing operation and the major-minor axis ratio screening, some scattered responses were removed, and the main crack area was retained, which can be used for subsequent parameter calculations.

[0052] Step 7: Crack Parameter Calculation

[0053] (1) Crack length calculation. Morphological ossification (also known as skeleton extraction) is performed on the crack region finally extracted in step six to obtain a crack skeleton image with a single pixel width (e.g., Figure 8 (As shown). Count the total number of skeleton pixels in the skeleton image. The pixel value of the crack length is... Combine the ratio of the physical size to the image pixel size obtained in step one. (In this embodiment) (per pixel), calculate the actual length of the crack:

[0054] (2) Calculation of average crack aperture. Calculate the pixel area of ​​the final crack region in step six. (i.e., the total number of pixels contained in the crack area), calculate the pixel value of the average crack aperture. Combined with the proportionality coefficient Converted to actual average opening:

[0055] (3) Crack surface density calculation. Crack surface density is defined as the ratio of the total area of ​​the crack region to the total area of ​​the image. The total area of ​​the image is... (in and (where the image height and width are in pixels, respectively), then the crack surface density is:

[0056] The calculated fracture parameters for the six core fracture images in this embodiment are shown in Table 1. Table 1 shows that the fracture lengths in the six images range from 5.4 cm to 13.2 cm, the average fracture aperture ranges from 0.14 cm to 0.27 cm, and the fracture surface density ranges from 0.0017 to 0.0070. These parameters reflect the differences in fracture length, aperture, and distribution area in different core images and can serve as a data basis for analyzing the degree of fracture development in cores.

[0057] Table 1 Crack parameters (Actual size: Image pixel size = 1:60)

[0058] Step 8: Explanation of Parameter Results

[0059] According to the calculation results in Table 1, there are certain differences in fracture length, average aperture, and areal density among the six core images. This result indicates that after fracture region extraction, this method can further provide parameters such as fracture length, average aperture, and areal density, facilitating unified statistical analysis of batches of core images.

[0060] In practical use, technicians can review the automatically extracted results. For images with clear crack boundaries and minimal background interference, the output parameters can be used directly. For images where cracks heavily overlap with rock textures, the threshold, neighborhood scale, or connected region filtering conditions can be adjusted based on the results of manual review. These methods can reduce the workload of manually drawing lines and measuring each image individually.

[0061] The processing time can be statistically analyzed based on the specific operating environment, image resolution, and parameter settings. In this embodiment, the algorithm can complete the extraction of fracture regions and parameter output for a single core image in a normal computer environment. In practical applications, the batch processing method can be determined based on the amount of on-site data and verification requirements.

[0062] Example 2

[0063] The difference between this embodiment and Embodiment 1 is that the core fracture image has a resolution of 800×600 pixels, is in TIFF format, and originates from a cored section of another oil and gas field. The ratio of the actual object size to the image pixel size is 1:40. This embodiment is used to illustrate the parameter adjustment methods under different image resolution conditions.

[0064] In the multi-scale local contrast calculation in step three, due to the reduced image resolution, this embodiment reduces the width of the outer region R2. The corresponding adjustments are 3, 7, and 11 (i.e.) =1, 3, 5), Gaussian kernel scale parameter The corresponding values ​​are adjusted to 10, 50, and 150. The remaining parameter settings are the same as in Example 1.

[0065] Example 3

[0066] This embodiment provides another set of parameter settings. In the image enhancement step, the scale number N of the multi-scale Gaussian function is set to 5, and the Gaussian kernel scale parameter... The weighting coefficients for each scale are 10, 40, 80, 160, and 250 respectively. All values ​​are set to 0.2. The local contrast calculation in step three uses four scales, corresponding to... Take 5, 9, 13 and 19 respectively (i.e. =2, 4, 6, 9). Weighting coefficients for crack saliency map fusion in step five. and Set the values ​​to 0.7 and 0.3 respectively. In step six, the radius of the structuring element for the morphological closing operation is set to 5 pixels, and the threshold for the ratio of the major axis to the minor axis is set to 4.

[0067] This embodiment illustrates that the Gaussian scale number, local contrast scale number, morphological structural element radius, and major-minor axis ratio threshold can be set according to image resolution and image quality, and are not limited to the specific parameters in Embodiment 1.

[0068] Example 4

[0069] The difference between this embodiment and Embodiment 1 is that, in step six, when removing scattered point regions, in addition to using the ratio of the major axis to the minor axis as a criterion, a screening condition for the area of ​​connected regions is added: connected regions with an area less than 50 pixels are directly identified as non-crack regions and removed. When there are many tiny bright spots or impurities on the core surface, a screening condition for the area of ​​connected regions can be added to remove smaller connected regions as noise regions.

[0070] This screening condition can be used in conjunction with the aspect ratio screening to reduce the impact of small noise areas on crack area statistics.

[0071] Example 5

[0072] The difference between this embodiment and Embodiment 1 lies in that, in step two, during image enhancement processing, the two enhanced images are fused using a weighted average method. This method has a relatively simple calculation process and can be used in processing scenarios where the calculation steps are relatively simplified. In images with weak crack edges or large differences in illumination, the covariance matrix fusion method in Embodiment 1 can also be used to retain more edge difference information. In practical applications, the appropriate fusion method can be selected according to the image quality and computational requirements.

[0073] This invention uses core fracture photographs as the processing object. After image enhancement and fracture region extraction, it further calculates fracture length, average aperture, and areal density through ossification and area statistics. This method can be used to uniformly process batches of core photographs, reducing the workload of manual image-by-image identification and statistics.

[0074] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for detecting core fracture parameters based on multi-scale local contrast measurement, characterized in that, Includes the following steps: Obtain the original image of the rock core crack, extract the scale bar from the image, and calculate the scaling factor between the actual size and the image pixel size; Image enhancement processing is performed on the original image to obtain an enhanced image with a more balanced brightness distribution; Based on the enhanced image, the local region contrast at multiple scales is calculated to obtain local contrast detection maps at each scale, and the local contrast detection maps at each scale are linearly weighted and fused to obtain a fused contrast map. Based on the enhanced image, the dark channel prior map is calculated using the dark channel prior principle; The fused contrast map and the dark channel prior map are linearly weighted and fused to extract the crack region and generate a crack saliency map. The crack saliency map is binarized to obtain a crack binary map; Morphological processing is performed on the binary image of the crack to identify and extract the crack region; Based on the extracted crack regions, calculate the crack length, average crack aperture, and crack surface density.

2. The method for detecting core fracture parameters based on multi-scale local contrast measurement according to claim 1, characterized in that, The image enhancement process includes: Convert the original image from the RGB color space to the HSV color space; The brightness channel components are extracted, and the illumination components are estimated based on the illumination reflection model and multi-scale Gaussian function. The brightness channel is enhanced according to the illumination component to obtain the enhanced brightness channel; The enhanced luminance channel is merged with the original saturation and chroma channels and converted back to the RGB color space.

3. The method for detecting core fracture parameters based on multi-scale local contrast measurement according to claim 2, characterized in that, The estimation of illumination components based on the illumination reflection model and multi-scale Gaussian function is implemented as follows: in, For the brightness channel of the input image, For the first Gaussian functions of various scales For the first The weight coefficients corresponding to each scale This represents the total number of scales.

4. The method for detecting core fracture parameters based on multi-scale local contrast measurement according to claim 1, characterized in that, The calculation of local region contrast at multiple scales specifically involves: For each pixel location in the enhanced image, an inner region R1 and an outer region R2 surrounding the inner region are constructed with that pixel location as the center. Calculate the mean eigenvectors of the inner region R1 and the outer region R2 in the CIE Lab color space, respectively. Calculate the Euclidean distance between the mean eigenvectors of the inner region R1 and the outer region R2, and use it as the local contrast value of the pixel at that scale. The above calculations were repeated at different external region scales to obtain local contrast detection maps at each scale.

5. The method for detecting core fracture parameters based on multi-scale local contrast measurement according to claim 4, characterized in that, The scale range of the external region R2 is: in, Let R2 be the scale parameter for the external region. and These are the height and width of the enhanced image, respectively.

6. The method for detecting core fracture parameters based on multi-scale local contrast measurement according to claim 1, characterized in that, The calculation of the dark channel prior map using the dark channel prior principle is implemented in the following way: in, For the input image in color channels The amount, In pixels A local image patch centered on the image.

7. The method for detecting core fracture parameters based on multi-scale local contrast measurement according to claim 1, characterized in that, The crack saliency map is binarized and segmented using the OTSU adaptive threshold segmentation algorithm.

8. The method for detecting core fracture parameters based on multi-scale local contrast measurement according to claim 1, characterized in that, The morphological processing of the binary crack image to identify and extract crack regions includes: Morphological closing operations are used to process the binary image of the crack to fill the voids inside the crack region. Calculate the ratio of the major axis length to the minor axis length of each connected region in the binary graph after the closing operation; Connected regions with a ratio less than a preset threshold are identified as non-crack regions and removed, while crack regions are retained.

9. The method for detecting core fracture parameters based on multi-scale local contrast measurement according to claim 1, characterized in that, The calculation of crack length specifically involves: performing morphological ossification on the extracted crack region to obtain a crack skeleton image, counting the total number of skeleton pixels in the crack skeleton image, and calculating the actual length of the crack by combining the scaling factor.

10. The method for detecting core fracture parameters based on multi-scale local contrast measurement according to claim 1, characterized in that, The average crack aperture and crack surface density are calculated as follows: Average crack aperture: The ratio of the extracted crack area to the crack length is calculated, and the actual average crack aperture is calculated by combining the ratio coefficient. Crack surface density: The ratio of the area of ​​the extracted crack region to the total area of ​​the enhanced image.

Images