A core multi-scale fracture extraction method based on fracture enhancement and screening technology

Abstract

The invention discloses a core multi-scale fracture extraction method based on fracture enhancement and screening technology, comprising: S1, acquiring an original CT scan image of a core containing fractures of different scales; S2, preprocessing the original CT scan image of the core to obtain a core denoised CT scan image and an information enhanced CT scan image; S3, performing macro-fracture segmentation based on the core denoised CT scan image to obtain a macro-fracture segmentation image, performing micro-fracture segmentation based on the information enhanced CT scan image to obtain a micro-fracture segmentation image, superimposing the macro-fracture segmentation image with the micro-fracture segmentation image to obtain a macro+micro-fracture segmentation image; S4, performing multiple fracture screening based on the macro+micro-fracture segmentation image to obtain a core multi-scale fracture extraction result, and adopting different segmentation methods for fractures of different scales, thereby improving the accuracy and stability of fracture identification.

Description

Technical Field

[0001] The present invention relates to the technical field of rock fracture identification, and more particularly to a core multi-scale fracture extraction method based on fracture enhancement and screening technology. Background Art

[0002] In the fields of geological exploration, oil and gas exploration and development, geotechnical engineering, etc., accurate identification of rock cracks is of great significance for the study of geological tectonic movement, evaluation of oil and gas reservoir and seepage space, evaluation of hydraulic fracture expansion effect, and prediction of slope stability. Traditional crack identification methods are mainly based on manual observation and analysis, which are inefficient and easily affected by subjective factors. In recent years, with the development of CT scanning and digital core technology, automated crack extraction methods have gradually become a research hotspot. However, existing crack extraction methods have the disadvantages of poor micro-crack identification effect, inability to accurately extract crack detail features, and difficulty in distinguishing cracks from pores. Especially for cores with coexisting cracks of different scales, the difference in crack opening is very large (up to more than 10 times), and it is difficult to accurately extract cracks using existing crack extraction methods. There is an urgent need to develop new crack extraction methods. Summary of the invention

[0003] In view of this, the purpose of the present invention is to provide a method for extracting multi-scale fractures in cores based on fracture enhancement and screening technology, so as to solve the shortcomings of the prior art, such as poor micro-fracture recognition effect, inability to accurately extract fracture detail features, and difficulty in distinguishing fractures from pores, thereby achieving accurate extraction of multi-scale fractures in cores.

[0004] To achieve the above object, the present invention provides the following technical solutions:

[0005] A core multi-scale fracture extraction method based on fracture enhancement and screening technology comprises the following steps:

[0006] S1, obtaining the original CT scan image of the core containing fractures of different scales;

[0007] S2. Preprocessing the original CT scan image of the core, including multi-scale crack filter kernel superposition noise reduction and micro-crack information enhancement, to obtain the core noise-reduced CT scan image and the information-enhanced CT scan image respectively;

[0008] S3, performing macro-crack segmentation based on the CT scan image after core noise reduction to obtain a macro-crack segmentation image, performing micro-crack segmentation based on the CT scan image after information enhancement to obtain a micro-crack segmentation image, and superimposing the macro-crack segmentation image with the micro-crack segmentation image to obtain a macro+micro-crack segmentation image;

[0009] S4. Multiple fracture screenings are performed based on the macro and micro fracture segmentation images to obtain the core multi-scale fracture extraction results.

[0010] Furthermore, in step S2, multi-scale crack filtering and superposition noise reduction specifically include the following steps:

[0011] a. Based on the original CT scan image of the core, read the minimum microcrack opening ε min and the maximum macro crack opening ε max , in the core fracture opening range (ε min , ε max ), extract n crack openings ε from small to large with equal spacing i , where (i=1, 2, ..., n);

[0012] b. Calculate the opening degree ε of n cracks i The corresponding filter kernel scale k i ;

[0013]

[0014] In the above formula: d is the original CT image resolution of the core, μm / pixel; ε i is the crack opening, μm; [] is the rounding operation; k i is the i-th filter kernel scale, i=1,2,…,n;

[0015] c. Use size k i ×k i The filter size of the original CT scan image of the core is r Filtering is performed and the average value of the images after filtering with different filter kernels is calculated to obtain the CT scan image I after core noise reduction. g ;

[0016]

[0017] In the above formula: I r is the original CT scan image of the core; k i is the scale of the i-th filter kernel; the position of the center pixel of the filter kernel is defined as the relative origin (0,0); (j,l) is the relative coordinate of any pixel in the filter kernel within the filter kernel; where, The Sort operator sorts the grayscale values ​​of the pixels at relative position (j, l) from small to large and outputs the grayscale values ​​of the pixels at the relative position (j, l). Pixel gray value; I g This is a CT scan image of the core after noise reduction.

[0018] Furthermore, in step S2, the microcrack information is enhanced, specifically:

[0019] Based on the CT scan image after core denoising I g, the micro-crack information is enhanced to obtain the CT scan image I after the core micro-crack information is enhanced e In formula (3), the square root part is responsible for the CT scan image I after core noise reduction g Perform crack boundary detection, and the part outside the square root is responsible for enhancing the information of microcracks;

[0020]

[0021] In the above formula: ε min is the minimum opening of microcracks, μm; [] is the rounding operation; (x, y) is the coordinate of the pixel point in the CT image; I e This is a CT scan image after core micro-fracture information enhancement.

[0022] Furthermore, the macro crack segmentation comprises the following steps:

[0023] a. Select the CT scan image after core noise reduction I g A region in the matrix that only contains macro cracks and matrix but no micro cracks is selected, and a grayscale histogram is drawn. The grayscale histogram shows a bimodal distribution, and the grayscale value of the lowest point between the two peaks is taken as the grayscale threshold thresh1 of the macro crack.

[0024] b. According to formula (4), the CT scan image I after core noise reduction is g Perform macro crack segmentation to obtain macro crack segmentation image I sMac ;

[0025]

[0026] Where: src is the image gray value matrix; thresh1 is the macro crack gray threshold; I sMac Macro crack segmentation image.

[0027] Furthermore, the microcrack segmentation comprises the following steps:

[0028] a. Extract CT scan image background I based on formula (5) f1 Based on the formula (6), the foreground I of the CT scan image is extracted f2 ;

[0029] b. Select CT scan image foreground I f2 The area that only contains microcracks and matrix but no macrocracks is plotted, and the grayscale histogram of the area is plotted. The grayscale histogram of the area will show a bimodal distribution, and the grayscale value of the lowest point between the two peaks is taken as the microcrack grayscale threshold thresh2;

[0030] c. Using formula (7) to calculate the foreground I of the CT scan image f2Perform binary segmentation to obtain the microcrack segmentation image I sMic ;

[0031]

[0032] In the above formula: is the contraction operation; is an expansion operation; [] is a rounding operation; I f1 It is the CT scan image background;

[0033] I f2 =I e -I f1 (6)

[0034] In the above formula: I f2 It is the foreground of the CT scan image;

[0035]

[0036] In the above formula: src is the image gray value matrix; thresh2 is the microcrack gray threshold; I sMic Microcrack segmentation image.

[0037] Furthermore, the results of macro crack segmentation and micro crack segmentation are superimposed, which specifically includes the following steps:

[0038] Formula (8) is used to segment the macro crack image I sMac and microcrack segmentation image I sMic Superimpose to obtain the macro + micro crack segmentation image I s ;

[0039]

[0040] In the above formula: I sMac is the macro crack segmentation image; I sMic I is the microcrack segmentation image; s This is the macro + micro crack segmentation image.

[0041] Furthermore, the multiple crack screenings include a first crack screening based on shape parameter analysis and a second crack screening based on flatness-elongation analysis.

[0042] Furthermore, the first crack screening based on shape parameter analysis is specifically as follows:

[0043] a. Calculate the macro + micro crack segmentation image I using formula (9) s The shape factor F of each region in the image is calculated, and the length L, width W, surface area S, volume V and shape factor F of each region are counted;

[0044]

[0045] In the above formula: N is the number of pixels on the surface of the target three-dimensional object; M is the total number of pixels of the target three-dimensional object; F is the shape factor;

[0046] b. According to the statistical data of length L, width W, surface area S, volume V and shape factor F, the length L-shape factor F cross-plot, width W-shape factor F cross-plot, surface area S-shape factor F cross-plot and volume V-shape factor F cross-plot are respectively prepared. The expressions of the linear fitting trend line of length L-shape factor F, the linear fitting trend line of width W-shape factor F, the linear fitting trend line of surface area S-shape factor F and the linear fitting trend line of volume V-shape factor F are respectively established by linear fitting as follows:

[0047]

[0048] In the above formula: L is length, μm; W is width, μm; S is surface area, μm 2 ; V is volume, μm 3 ; F is the shape factor; a1, b1, a2, b2, a3, b3, a4, b4 are fitting coefficients;

[0049] c. In the cross-plot of length L-shape factor F, the cross-plot of width W-shape factor F, the cross-plot of surface area S-shape factor F, and the cross-plot of volume V-shape factor F, two screening boundary lines parallel to the fitting trend line are drawn, and the expressions are shown in Formulas 11 to 14. By adjusting the parameters c1, c2, c3, and c4, the number of data points in the screening area surrounded by the two screening boundary lines accounts for 80% of the total number of data points. The obtained screening data A, B, C, and D are the crack area data predicted by the cross-plot of length L-shape factor F, the cross-plot of width W-shape factor F, the cross-plot of surface area S-shape factor F, and the cross-plot of volume V-shape factor F, respectively;

[0050]

[0051] In the above formula: c1, c2, c3, c4 are the fitting trend line movement parameters; A, B, C, D are the crack area data predicted by the length L ~ shape factor F cross plot, width W ~ shape factor F cross plot, surface area S ~ shape factor F cross plot, volume V ~ shape factor F cross plot, respectively;

[0052] d. Obtain the intersection of the screening data A, B, C, and D according to formula (15) to obtain the first screening result COMB1;

[0053] COMB1=A∩B∩C∩D (15)

[0054] In the above formula: ∩ is the intersection operator; COMB1 is the first screening result.

[0055] Furthermore, the second crack screening based on flatness-elongation analysis is:

[0056] a. Calculate the flatness and elongation of each area after the first screening according to formula (16) and formula (17);

[0057]

[0058] In the above formula: S is the minimum eigenvalue of the image covariance matrix; M is the middle eigenvalue; Flatness is the flatness;

[0059]

[0060] In the above formula: L is the maximum eigenvalue of the image covariance matrix; Elongation is the elongation;

[0061] b. Substitute the flatness and elongation of each area in the first screening result COMB1 into formula (18) to obtain the second screening result COMB2;

[0062]

[0063] In the above formula: Elongation is elongation; Flatness is flatness; COMB2 is the second screening result;

[0064] The second screening result COMB2 obtained by flatness-elongation analysis removes the mineral boundary surface area with small differences in fracture shape characteristics. The second screening result COMB2 is the final result of multi-scale fracture extraction of the core.

[0065] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0066] 1. The proposed microcrack information enhancement technology effectively highlights and strengthens microcracks; 2. Different segmentation methods are used for cracks of different scales to improve the accuracy and stability of crack identification; 3. Through the relationship between crack shape factor and length, width, surface area and volume, elongation and flatness, accurate distinction between cracks and non-cracks is completed. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0068] The core multi-scale fracture extraction method of the fracture enhancement and screening technology of the present invention is further described below in conjunction with the accompanying drawings;

[0069] Figure 1 It is a flow chart of a core multi-scale fracture extraction method based on fracture enhancement and screening technology provided by the present invention;

[0070] Figure 2 It is a schematic diagram of the principle of multi-scale crack filter kernel superposition noise reduction provided by the present invention;

[0071] Figure 3 It is a comparison diagram of images before and after microcrack information enhancement provided by the present invention; wherein, Figure 3 (a) is the image before microcrack information enhancement; Figure 3 (b) is the image after microcrack information enhancement;

[0072] Figure 4 is a schematic diagram of the macro crack segmentation principle provided by the present invention; wherein, Figure 4 (a) is the CT scan image after core noise reduction; Figure 4 (b) is the grayscale histogram; Figure 4 (c) is the macro crack segmentation image;

[0073] Figure 5 Schematic diagram of the microcrack segmentation principle provided by the present invention; wherein, Figure 5 (a) is the CT scan image after microcrack information enhancement; Figure 5 (b) is the foreground of the CT scan image; Figure 5 (c) is the CT scan image background; Figure 5 (d) is the grayscale histogram; Figure 5 (e) is the microcrack segmentation image;

[0074] Figure 6 : is a schematic diagram of the segmentation result superposition principle provided by the present invention; wherein, Figure 6 (a) is the macro crack segmentation image; Figure 6 (b) is the microcrack segmentation image; Figure 6 (c) is the macro + micro crack segmentation image;

[0075] Figure 7 is a schematic diagram of shape parameter analysis of each region in the core image provided by the present invention; wherein, Figure 7(a) Reconstruction of each region in the core image; Figure 7 (b) is the cross-plot of length L and shape factor F of each region; Figure 7 (c) is the cross-plot of width W and shape factor F of each region; Figure 7 (d) is the cross-plot of surface area S and shape factor F in each region; Figure 7 (e) is the cross-plot of volume V and shape factor F in each region;

[0076] Figure 8 It is a schematic diagram of fracture-non-fracture flatness-elongation in the core multi-scale fracture extraction method based on fracture enhancement and screening technology provided by the present invention; wherein, Figure 8 (a) is the crack-non-crack theoretical shape diagram; Figure 8 (b) is the crack-non-crack theoretical shape flatness-elongation distribution;

[0077] Fig. 9 It is a schematic diagram of a micro plunger core sample photo and a CT scan image in the core multi-scale fracture extraction method based on fracture enhancement and screening technology provided by the present invention; wherein, Fig. 9 (a) is a core photo; Fig. 9 (b) 3D reconstruction of CT scan (0.3 μm / pixel); Fig. 9 (c) is an orthogonal slice of a CT scan;

[0078] Fig.10 It is a filter noise reduction optimization effect diagram of the core sample in the core multi-scale fracture extraction method based on fracture enhancement and screening technology provided by the present invention using the filter kernel 3, 7, 15 combination; wherein, Fig.10 (a) is a two-dimensional slice of the core sample;

[0079] Fig.11 It is a diagram of the effect of enhancing the information of core samples in the core multi-scale fracture extraction method based on the fracture enhancement and screening technology provided by the present invention; wherein, Fig.11 (a) is a two-dimensional slice of the core sample; Fig.11 (b) is the enlargement of area I; Fig.11 (c) is the crack information enhancement effect in region I; Fig.11 (d) is the enlarged view of region II; Fig.11 (e) Crack information enhancement effect in region II;

[0080] Fig.12 It is a diagram showing the effect of segmenting fractures of different scales in a core sample in the core multi-scale fracture extraction method based on fracture enhancement and screening technology provided by the present invention; wherein, Fig.12 (a) is a three-dimensional image of the core sample; Fig.12 (b) is a two-dimensional slice; Fig.12 (c) is a macro crack; Fig.12(d) micro cracks; Fig.12 (e) non-crack;

[0081] Fig.13 It is a diagram showing the effect of analyzing the shape parameters of each region of a core sample in the core multi-scale fracture extraction method based on fracture enhancement and screening technology provided by the present invention; wherein, Fig.13 (a) Reconstruction of each region in the core image; Fig.13 (b) is the cross-plot of length L and shape factor F of each region; Fig.13 (c) is the cross-plot of width W and shape factor F of each region; Fig.13 (d) is the cross-plot of surface area S and shape factor F in each region; Fig.13 (e) is the cross-plot of volume V and shape factor F in each region;

[0082] Fig.14 It is a crack-non-crack flatness-elongation analysis effect diagram of a core sample in a core multi-scale crack extraction method based on crack enhancement and screening technology provided by the present invention; wherein, Fig.14 (a) is the fracture-non-fracture shape diagram of the core sample; Fig.14 (b) is the fracture-non-fracture flatness-elongation distribution of the core sample;

[0083] Fig.15 It is a result diagram of core sample fracture extraction in the core multi-scale fracture extraction method based on fracture enhancement and screening technology provided by the present invention; wherein, Fig.15 (a) is a 3D reconstruction of a CT scan; Fig.15 (b) is CT orthogonal slice; Fig.15 (c) is an orthogonal slice of the crack; Fig.15 (d) is the crack classification. DETAILED DESCRIPTION

[0084] The specific implementation of the present invention is further described in detail below in conjunction with the accompanying drawings and examples. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0085] In order to better understand the purpose, structure and function of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings.

[0086] like Figure 1 As shown, the present invention provides a core multi-scale fracture extraction method based on fracture enhancement and screening technology, and the detailed principle and operation of the following four steps are described as follows:

[0087] (1) Image data preparation: Prepare the original CT scan image of the core with fractures of different scales. rThis method requires that the original CT image resolution d of the core is higher than 50 μm / pixel (i.e., CT image resolution d < 50 μm / pixel). If the CT scanning equipment allows, scanning the core with a higher image resolution can further improve the accuracy of fracture extraction.

[0088] (2) Enhancement of crack detail information: ① Multi-scale crack filter kernel superposition denoising

[0089] The multi-scale fracture filter kernel superposition denoising method is used to denoise the original CT scan image of the core to remove noise information such as salt and pepper noise and impulse noise. The multi-scale fracture filter kernel superposition denoising method includes three steps:

[0090] a. Observe the original CT scan image of the core and read the minimum opening of the microcrack ε min and the maximum macro crack opening ε max , in the core fracture opening range (ε min , ε max ), extract n crack openings ε from small to large with equal spacing i (i=1, 2,…, n).

[0091] b. Calculate the opening degree ε of n cracks i The corresponding filter kernel scale k i , the specific calculation method is as follows:

[0092]

[0093] Where: d is the original CT image resolution of the core (μm / pixel); ε i is the crack opening (μm); [] is the rounding operation; k i is the i-th filter kernel scale (pixel), i=1,2,…,n (usually, n=3 can meet the requirements of core fracture extraction).

[0094] c. Figure 2 As shown, the size is k i ×k i The filter size of the original CT scan image of the core is r Filtering is performed, and the average value of the images after filtering with multiple different filter kernels is calculated to obtain the CT scan image I after core noise reduction. g , the specific calculation method is as follows:

[0095]

[0096] In the above formula: I r is the original CT scan image of the core; k iis the scale of the ith filter kernel; the position of the center pixel of the filter kernel is defined as the relative origin (0,0); (j,l) is the relative coordinate of any pixel in the filter kernel within the filter kernel, The Sort operator sorts the grayscale values ​​of the pixels at relative position (j, l) from small to large and outputs the grayscale values ​​of the pixels at the relative position (j, l). Pixel gray value; I g This is a CT scan image of the core after noise reduction.

[0097] ② Microcrack information enhancement

[0098] like Figure 3 As shown in the figure, in the CT scan image after core noise reduction, the grayscale values ​​of microcracks and matrix are still similar, so it is necessary to enhance the microcrack information so as to extract microcracks more accurately in the future.

[0099] CT scan image after core noise reduction I g , use formula (3) to perform microcrack information enhancement processing, and obtain the CT scan image I after core microcrack information enhancement e In formula (3), the square root part is responsible for the CT scan image I after core noise reduction. g The first part detects the crack boundary, and the other part is responsible for enhancing the information of microcracks. Microcrack information enhancement technology can effectively highlight and strengthen microcracks, laying the foundation for accurate crack segmentation and extraction. Figure 3 shown.

[0100]

[0101] In the above formula: ε min is the minimum opening of the microcrack (μm); [] is the rounding operation; (x, y) is the pixel coordinate of the CT image; I e This is a CT scan image after core micro-fracture information enhancement.

[0102] (3) Differentiated crack segmentation

[0103] ① Macro crack segmentation

[0104] like Figure 4 As shown in Figure 2, macro crack segmentation consists of two steps:

[0105] a. Select the CT scan image after core noise reduction I g A region in the matrix that contains only macro cracks and matrix but no micro cracks, such as Figure 4 (a) shows; draw the grayscale histogram of the area, as shown in Figure 4 (b) shows that the grayscale histogram of this area will show a bimodal distribution, and the grayscale value of the lowest point between the two peaks is taken as the macro crack grayscale threshold thresh1, as shown in Figure 4(b) as shown.

[0106] b. According to formula (4), the CT scan image I after core noise reduction is g Perform macro crack segmentation to obtain macro crack segmentation image I sMac ,like Figure 4 (c) shows the macro crack segmentation image I sMac In the figure, the pixel value of 1 contains complete macro cracks and a small number of larger pores.

[0107]

[0108] In the above formula: src is the image gray value matrix; thresh1 is the macro crack gray threshold (dimensionless); I sMac Macro crack segmentation image.

[0109] ② Micro crack segmentation

[0110] Microfracture segmentation in CT scan image after core microfracture information enhancement I e Based on Figure 5 As shown in (a); microcrack segmentation includes three steps:

[0111] a. Using formula (5) to extract the CT scan image background I f1 On the basis of Figure 5 (c) As shown in the figure, the foreground I of the CT scan image is extracted using formula (6) f2 ,like Figure 5 (b) as shown.

[0112] b. Select CT scan image foreground I f2 A region in the matrix that contains only microcracks and matrix but no macrocracks, such as Figure 5 (b) shows; draw the grayscale histogram of the area, as shown in Figure 5 (d) As shown; the grayscale histogram of this area will show a bimodal distribution, and the grayscale value of the lowest point between the two peaks is taken as the microcrack grayscale threshold thresh2, as shown in Figure 5 (d) as shown.

[0113] c. Using formula (7) to calculate the foreground I of the CT scan image f2 Perform binary segmentation to obtain the microcrack segmentation image I sMic ,like Figure 5 (e) shows the microcrack segmentation image I sMic In the figure, the pixel points with a pixel value of 1 include complete microcracks, some incomplete macrocracks, and a small number of special mineral boundaries.

[0114]

[0115] Where: is the contraction operation; is an expansion operation; [] is a rounding operation; I f1 It is the CT scan image background;

[0116] I f2 =I e -I f1 (6)

[0117] In the above formula: I f2 Foreground of CT scan image.

[0118]

[0119] Where: src is the image gray value matrix; thresh2 is the microcrack gray threshold (dimensionless); I sMic Microcrack segmentation image.

[0120] ③Segmentation result superposition

[0121] The macro crack segmentation image I obtained by differential crack segmentation sMac and microcrack segmentation image I sMic Superposition (Formula 8) is performed to obtain the macro + micro crack segmentation image I s ( Figure 6 ). Due to the macro crack segmentation image I sMac The macro cracks in the image completely contain the micro cracks segmented image I sMic The incomplete macro cracks in the sMac with I sMic Superimposed macro and micro crack segmentation image I s In the image, the pixel value of 1 includes complete microcracks, complete macrocracks, a small number of large pores and a small number of special mineral boundaries; the macro + microcrack segmentation image I s Perform crack screening, eliminate non-cracks, and obtain more accurate crack extraction results.

[0122]

[0123] In the above formula: I sMac is the macro crack segmentation image; I sMic I is the microcrack segmentation image; s This is the macro + micro crack segmentation image.

[0124] (4) Multiple crack screening

[0125] ① Shape parameter analysis for the first crack screening

[0126] Shape parameter analysis for first crack screening in macro + micro crack segmentation image I s Based on this, there are 4 steps:

[0127] a. Use formula (9) to segment the macro + micro crack image I s The shape factor F is calculated for each individual area extracted (including macro cracks, micro cracks, a small number of larger pores, and a small number of special mineral boundaries), and the length L, width W, surface area S, volume V and shape factor F of each area are counted.

[0128]

[0129] Where: N is the number of pixels on the surface of the target three-dimensional object; M is the total number of pixels of the target three-dimensional object; F is the shape factor (dimensionless).

[0130] b. According to the statistical length L, width W, surface area S, volume V and shape factor F data, L-F intersection diagram (length-shape factor intersection diagram), W-F intersection diagram (width-shape factor intersection diagram), S-F intersection diagram (surface area-shape factor intersection diagram), V-F intersection diagram (volume-shape factor intersection diagram) are respectively prepared, such as Figure 7 As shown; through linear fitting, the L~F linear fitting trend line, W~F linear fitting trend line, S~F linear fitting trend line, and V~F linear fitting trend line are established respectively, and the expression of the linear fitting trend line is formula (10).

[0131]

[0132] Where: L is length (μm); W is width (μm); S is surface area (μm 2 ); V is volume (μm 3 ); F is the shape factor (dimensionless); a1, b1, a2, b2, a3, b3, a4, b4 are fitting coefficients.

[0133] c. In the L-F crossplot, W-F crossplot, S-F crossplot, and V-F crossplot, draw two straight lines (screening boundary lines) parallel to the fitting trend line, and adjust the parameters c1, c2, c3, and c4 so that the number of data points in the screening area surrounded by the two screening boundary lines accounts for 80% of the total number of data points (Formula 11 to Formula 14). The obtained screening data A, B, C, and D are the crack area data predicted by the L-F crossplot, W-F crossplot, S-F crossplot, and V-F crossplot, respectively. Figure 7 shown.

[0134]

[0135] In the above formula, c1, c2, c3, and c4 are the moving parameters of the fitting trend line; A, B, C, and D are the crack area data predicted by the L-F crossplot, W-F crossplot, S-F crossplot, and V-F crossplot, respectively.

[0136] d. According to formula (15), the intersection of screening data A, B, C, and D is obtained to obtain the first screening result COMB1; the first crack screening can remove the macro + micro crack segmentation image I by shape parameter analysis. s The larger pore areas with different shape characteristics from the cracks are selected to achieve the initial screening of cracks.

[0137] COMB1=A∩B∩C∩D (15)

[0138] In the above formula: ∩ is the intersection operator; COMB1 is the first screening result.

[0139] ② Flatness-elongation analysis for the second crack screening

[0140] The second crack screening by flatness-elongation analysis is carried out based on the first screening result COMB1, and is divided into the following two steps:

[0141] a. Calculate the flatness and elongation of each area after the first screening according to formula (16) and formula (17).

[0142]

[0143] Where: S is the minimum eigenvalue of the image covariance matrix (dimensionless); M is the intermediate eigenvalue (dimensionless); Flatness is the flatness (dimensionless).

[0144]

[0145] Where: L is the maximum eigenvalue of the image covariance matrix (dimensionless); Elongation is the elongation (dimensionless).

[0146] b. Substitute the flatness and elongation of each area in the first screening result COMB1 into formula (18) to obtain the second screening result COMB2.

[0147]

[0148] Wherein: Elongation is elongation (dimensionless); Flatness is flatness (dimensionless); COMB2 is the second screening result.

[0149] The second screening result COMB2 obtained through flatness-elongation analysis can remove special mineral boundary surface areas with small differences in fracture shape characteristics. The second screening result COMB2 is the final result of multi-scale fracture extraction of cores.

[0150] The present invention further adopts the following specific implementation methods based on the collected micro core plunger samples as an example:

[0151] (1) Preparing image data

[0152] Prepare a core sample with a diameter of 1.2 mm and a length of 2.0 mm, and the lithology is feldspar lithic sandstone, such as Figure 8 As shown in (a), the core sample has pores, macro cracks and micro cracks. The occurrence of macro cracks and micro cracks is mainly high angle. There is no filling phenomenon in the cracks. Some macro cracks are in contact with pores. According to the crack opening, the core cracks can be divided into two categories: macro cracks with an opening greater than 18μm and micro cracks with an opening less than 8μm. The opening of the largest macro crack is ε max The minimum microcrack opening is 23μm and the minimum microcrack opening is ε min The original CT image resolution d of the core sample is 0.3 μm / pixel. The grayscale values ​​of the images of different types of cracks and matrix in the sample are as follows from high to low: matrix, microcracks, macrocracks and pores. Fig. 9 (b) and Fig. 9 (c) The difficulties in extracting fractures from core samples are as follows: ① The grayscale values ​​of matrix and micro-fractures are very close and difficult to distinguish; ② The grayscale values ​​of macro-fractures and pores are basically the same and cannot be directly distinguished by threshold.

[0153] (2) Crack detail information enhancement

[0154] ①Multi-scale crack filter kernel superposition noise reduction

[0155] a. From the prepared image data, we can see that the core fracture opening range is (5μm, 23μm); extract three fracture openings ε from (5μm, 23μm) with equal intervals from small to large i , respectively ε1=5μm, ε2=14μm, ε3=23μm.

[0156] b. The crack opening ε i Substitute the original CT image resolution d (0.3 μm pixel) of the core into formula (1) to calculate the filter kernel scale k corresponding to the three opening degrees. i (i=1, 2, 3) are 3, 7, and 15 respectively.

[0157] c. Set the filter kernel scale ki and the original CT scan image of the core I r Substitute into formula (2) to filter the image and obtain the CT scan image I after core noise reduction: g ,like Fig.10 As shown; through Fig.10 (b1), Fig.10 (b2), Fig.10 (b3) Superposition and averaging yield Fig.10 (d1), Fig.10 (c1), Fig.10 (c2), Fig.10 (c3) Superposition and averaging give Fig.10 (d2).

[0158] ② Microcrack information enhancement

[0159] CT scan image after core noise reduction I g 、The minimum opening of microcracks ε min Substituting the original CT image resolution d (5 μm) and the core into formula (3), the CT scan image I after the core microcrack information is enhanced is obtained. e ,like Fig.11 shown.

[0160] (3) Differentiated crack segmentation

[0161] ① Macro crack segmentation: a. Select the CT scan image after core noise reduction I g For a region that only contains macro cracks and matrix, the gray value of the lowest point between two peaks in its gray histogram is read out as the macro crack gray threshold thresh1 (thresh1=20).

[0162] b. According to formula (4), the CT scan image I after core noise reduction is g Perform macro crack segmentation to obtain macro crack segmentation image I sMac ,like Fig.12 As shown in (c1).

[0163] ② Micro-fracture segmentation: a. CT scan image after enhancing the micro-fracture information of the core I e 、The minimum opening of microcracks ε min (5 μm) and the original CT image resolution d (0.3 μm / pixel) of the core are substituted into formula (5) to obtain the CT scan image background I f1 CT scan image enhanced with core microcrack information I e Subtract CT scan image background I f1 (Formula 6), we get the CT scan image foreground I f2 .

[0164] b. Extracting CT scan image foreground I f2 For a region that only contains microcracks and matrix, the gray value of the lowest point between two peaks in its gray histogram is read out as the microcrack gray threshold thresh2 (thresh2=65).

[0165] c. According to formula (7), the CT scan image I after the core microcrack information is enhanced e Perform microcrack segmentation to obtain the microcrack segmentation image I sMic ,like Fig.12 (d1) as shown.

[0166] ③Segmentation result superposition

[0167] The macro crack segmentation image I obtained by differential crack segmentation sMac and microcrack segmentation image I sMic The macro + micro crack segmentation image I is calculated by formula (8): s .

[0168] (4) Multiple crack screening

[0169] ① Shape parameter analysis for the first crack screening: a. Statistics I s The parameters of each region in the image are length L, width W, surface area S and volume V; the macro + micro crack segmentation image I is calculated using formula (9) s The shape factor F of each region in .

[0170] b. According to the statistical length L, width W, surface area S, volume V and shape factor F data, L-F intersection diagram, W-F intersection diagram, S-F intersection diagram and V-F intersection diagram are respectively prepared, such as Fig.13 As shown; through linear fitting, the L~F linear fitting trend line, W~F linear fitting trend line, S~F linear fitting trend line, and V~F linear fitting trend line are established respectively, and the expression of the linear fitting trend line is formula (19).

[0171]

[0172] c. Fig.13 As shown, according to formulas (11) to (14), the fitting trend line moving parameters c1, c2, c3, and c4 are adjusted so that the number of data points in the screening area accounts for 80% of the total number of data points (formulas 20 to 23), and the screening data A, B, C, and D are obtained.

[0173]

[0174] d. Then substitute the screening data A, B, C, and D into formula (15) to obtain the first screening result COMB1.

[0175] ② Flatness-elongation analysis for the second crack screening: a. After the shape parameter analysis, the first crack screening is performed to obtain the first screening result COMB1. The length L, width W, surface area S, volume V and shape factor F of each area in the first screening result COMB1 are shown in Table 1. The elongation Flatness and flatness Elongation of each area in the first screening result COMB1 are calculated by formulas (9) and (10) (as shown in Table 1).

[0176] Table 1 Parameters of each region in the first screening results

[0177]

[0178]

[0179] b. Substitute the flatness and elongation of each area in the first screening result COMB1 into formula (18) to obtain the second screening result COMB2. The screening analysis results are as follows: Fig.14 As shown in the figure; after fracture detail information enhancement, differentiated fracture segmentation and multiple fracture screening, the fracture area in the micro core plunger sample of a certain area was successfully extracted, as shown in the figure. Fig.15 shown.

[0180] The meanings of the various parameters in the present invention are as follows:

[0181] Parameter Table

[0182]

[0183]

[0184] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A core multi-scale fracture extraction method based on fracture enhancement and screening technology, characterized in that: The following steps are involved: S1, obtaining the original CT scan image of the core containing fractures of different scales; S2. Preprocessing the original CT scan image of the core, including multi-scale crack filter kernel superposition noise reduction and micro-crack information enhancement, to obtain the core noise-reduced CT scan image and the information-enhanced CT scan image, respectively, specifically including the following steps: a. Based on the original CT scan image of the core, read the minimum microcrack opening ε min and the maximum macro crack opening ε max , in the core fracture opening range (ε min , ε max ), extract n crack openings ε from small to large with equal spacing i , where (i=1, 2, ..., n); b. Calculate the opening degree ε of n cracks i The corresponding filter kernel scale k i ; In the above formula: d is the original CT image resolution of the core, μm / pixel; ε i is the crack opening, μm; [] is the rounding operation; k i is the i-th filter kernel scale, i=1,2,…,n; c. Use size k i ×k i The filter size of the original CT scan image of the core is r Filtering is performed and the average value of the images after filtering with different filter kernels is calculated to obtain the CT scan image I after core noise reduction. g ; In the above formula: I r is the original CT scan image of the core; k i is the scale of the i-th filter kernel; the position of the center pixel of the filter kernel is defined as the relative origin (0,0); (j,l) is the relative coordinate of any pixel in the filter kernel within the filter kernel; where, The Sort operator sorts the grayscale values ​​of the pixels at relative position (j, l) from small to large and outputs the grayscale values ​​of the pixels at the relative position (j, l). Pixel gray value; I g This is the CT scan image after core noise reduction; In S2, microcrack information is enhanced, specifically: Based on the CT scan image after core denoising I g , the micro-crack information is enhanced to obtain the CT scan image I after the core micro-crack information is enhanced e In formula (3), the square root part is responsible for the CT scan image I after core noise reduction g Perform crack boundary detection, and the part outside the square root is responsible for enhancing the information of microcracks; In the above formula: ε min is the minimum opening of microcracks, μm; [] is the rounding operation; (x, y) is the coordinate of the pixel point in the CT image; I e This is the CT scan image after the core micro-fracture information is enhanced; S3, performing macro-crack segmentation based on the CT scan image after core noise reduction to obtain a macro-crack segmentation image, performing micro-crack segmentation based on the CT scan image after information enhancement to obtain a micro-crack segmentation image, and superimposing the macro-crack segmentation image with the micro-crack segmentation image to obtain a macro+micro-crack segmentation image, specifically comprising the following steps: The macro crack segmentation image I is obtained by using formula (4). sMac and microcrack segmentation image I sMic Superimpose to obtain the macro + micro crack segmentation image I s ; In the above formula: I sMac is the macro crack segmentation image; I sMic I is the microcrack segmentation image; s It is the macro + micro crack segmentation image; S4. Perform multiple fracture screening based on the macro + micro fracture segmentation image to obtain the core multi-scale fracture extraction results; wherein the multiple fracture screening includes: a first fracture screening based on shape parameter analysis and a second fracture screening based on flatness-elongation analysis; The first crack screening based on shape parameter analysis is specifically as follows: a. Calculate the macro + micro crack segmentation image I using formula (5) s The shape factor F of each region in the image is calculated, and the length L, width W, surface area S, volume V and shape factor F of each region are counted; In the above formula: N is the number of pixels on the surface of the target three-dimensional object; M is the total number of pixels of the target three-dimensional object; F is the shape factor; b. According to the statistical data of length L, width W, surface area S, volume V and shape factor F, the length L-shape factor F cross-plot, width W-shape factor F cross-plot, surface area S-shape factor F cross-plot and volume V-shape factor F cross-plot are respectively prepared. The expressions of the linear fitting trend line of length L-shape factor F, the linear fitting trend line of width W-shape factor F, the linear fitting trend line of surface area S-shape factor F and the linear fitting trend line of volume V-shape factor F are respectively established by linear fitting as follows: In the above formula: L is length, μm; W is width, μm; S is surface area, μm 2 ; V is volume, μm 3 ; F is the shape factor; a1, b1, a2, b2, a3, b3, a4, b4 are fitting coefficients; c. In the cross-plot of length L-shape factor F, the cross-plot of width W-shape factor F, the cross-plot of surface area S-shape factor F, and the cross-plot of volume V-shape factor F, two screening boundary lines parallel to the fitting trend line are drawn, and the expressions are shown in Formulas 11 to 14. By adjusting the parameters c1, c2, c3, and c4, the number of data points in the screening area surrounded by the two screening boundary lines accounts for 80% of the total number of data points. The obtained screening data A, B, C, and D are the crack area data predicted by the cross-plot of length L-shape factor F, the cross-plot of width W-shape factor F, the cross-plot of surface area S-shape factor F, and the cross-plot of volume V-shape factor F, respectively; In the above formula: c1, c2, c3, c4 are the fitting trend line movement parameters; A, B, C, D are the crack area data predicted by the length L ~ shape factor F cross plot, width W ~ shape factor F cross plot, surface area S ~ shape factor F cross plot, volume V ~ shape factor F cross plot, respectively; d. Obtain the intersection of the screening data A, B, C, and D according to formula (11) to obtain the first screening result COMB1; COMB1=A∩B∩C∩D (11) In the above formula: ∩ is the intersection operator; COMB1 is the first screening result; The second crack screening based on flatness-elongation analysis is specifically as follows: a. Calculate the flatness and elongation of each area after the first screening according to formula (12) and formula (13); In the above formula: S is the minimum eigenvalue of the image covariance matrix; M is the middle eigenvalue; Flatness is the flatness; In the above formula: L is the maximum eigenvalue of the image covariance matrix; Elongation is the elongation; b. Substitute the flatness and elongation of each area in the first screening result COMB1 into formula (14) to obtain the second screening result COMB2; In the above formula: Elongation is elongation; Flatness is flatness; COMB2 is the second screening result; The second screening result COMB2 obtained by flatness-elongation analysis removes the mineral boundary surface area with small differences in fracture shape characteristics. The second screening result COMB2 is the final result of multi-scale fracture extraction of the core.

2. The core multi-scale fracture extraction method based on fracture enhancement and screening technology according to claim 1 is characterized in that: The macro crack segmentation comprises the following steps: a. Select the CT scan image after core noise reduction I g A region in the matrix that only contains macro cracks and matrix but no micro cracks is selected, and a grayscale histogram is drawn. The grayscale histogram shows a bimodal distribution, and the grayscale value of the lowest point between the two peaks is taken as the grayscale threshold thresh1 of the macro crack. b. According to formula (15), the CT scan image I after core noise reduction is g Perform macro crack segmentation to obtain macro crack segmentation image I sMac ; Where: src is the image gray value matrix; thresh1 is the macro crack gray threshold; I sMac Macro crack segmentation image.

3. The core multi-scale fracture extraction method based on fracture enhancement and screening technology according to claim 1 is characterized in that: The microcrack segmentation comprises the following steps: a. Extract CT scan image background I based on formula (16) f1 Based on the formula (17), the foreground I of the CT scan image is extracted f2 ; b. Select CT scan image foreground I f2 The area that only contains microcracks and matrix but no macrocracks is plotted, and the grayscale histogram of the area is plotted. The grayscale histogram of the area will show a bimodal distribution, and the grayscale value of the lowest point between the two peaks is taken as the microcrack grayscale threshold thresh2; c. Using formula (18) to calculate the foreground I of the CT scan image f2 Perform binary segmentation to obtain the microcrack segmentation image I sMic ; In the above formula: is the contraction operation; is an expansion operation; [] is a rounding operation; I f1 It is the CT scan image background; I f2 =I e -I f1 (17) In the above formula: I f2 It is the foreground of the CT scan image; In the above formula: src is the image gray value matrix; thresh2 is the microcrack gray threshold; I sMic Microcrack segmentation image.

Images