A method for quickly obtaining structural plane parameters in a borehole

Abstract

The application discloses a method for quickly obtaining structural plane parameters in a borehole, and belongs to the technical field of intelligent rapid detection of surrounding rock fissures in geotechnical engineering and mining engineering. A neural network is used to establish a structural plane pixel identification model; the structural plane pixels of an image in an anchoring hole are identified by using the structural plane pixel identification model, and a thinning algorithm is used to reduce the number of structural plane pixels; structural plane skeleton pixels and other pixels of the image in the anchoring hole are labeled, each column of structural plane skeleton pixels is segmented, and thus the continuous structural plane skeleton pixels in each column are divided into a segment; a structural plane skeleton expression is obtained; and structural plane parameters, including a structural plane position, an inclination angle and a tendency, are output according to the structural plane skeleton expression. The method can solve the problems of slow speed and poor precision in the detection of surrounding rock fissures on site, and effectively improves the intelligent rapid identification efficiency of structural planes by using the application, thereby providing an important reference for safe construction of geotechnical engineering, mining engineering and the like.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent detection of surrounding rock cracks, and particularly relates to a method for quickly obtaining structural plane parameters in a borehole. BACKGROUND

[0002] Geotechnical engineering, bridge and tunnel engineering and the like have high requirements for geological exploration, and pay particular attention to crack distribution characteristics, which have an important influence on safe construction. Borehole imaging technology has been widely applied to the field of surrounding rock detection, and the borehole wall rock characteristics can be clearly seen through the anchoring hole image. However, the analysis of rock characteristics mainly relies on the experience of professional personnel, is greatly influenced by subjectivity, and some small cracks may be ignored, and manual surrounding rock analysis is slow, so an intelligent surrounding rock identification system is urgently needed to intelligently and quickly analyze the characteristics of surrounding rock and ensure safe construction.

[0003] The prior art CN113298140A discloses a method for quantitatively evaluating and classifying mining rock mass quality based on a borehole image, and relates to the technical field of mining engineering and solid material measurement. The steps of the method include: drilling a borehole in the surrounding rock of a roadway before the influence of mining, and observing the borehole image of the surrounding rock structural plane; dividing the filling closed structural plane, the filling cracked structural plane, the transverse cracked structural plane, the longitudinal cracked structural plane and the transverse broken zone; classifying and counting the structural plane on the borehole image to determine the comprehensive width of the structural plane in each evaluation area; observing the borehole image of the surrounding rock structural plane again after the borehole is affected by mining, and dividing the structural plane category, classifying and counting to determine the comprehensive width of the structural plane in each evaluation area; comparing the maximum value of the comprehensive width of the structural plane, classifying the quality of the mining rock mass, and taking measures to reinforce the surrounding rock of each classified rock mass. The method can only provide a general classification, and the precision is low. SUMMARY

[0004] In view of the deficiencies of the prior art, a method for quickly obtaining structural plane parameters in a borehole is provided, which has simple steps, is convenient to use, and can quickly obtain the structural plane characteristics in the borehole.

[0005] To achieve the above technical purpose, the present application discloses a method for quickly obtaining structural plane parameters in a borehole, and the specific steps are as follows:

[0006] S100, a large number of in-situ anchoring hole images are collected, a structural plane pixel recognition model is established using a neural network, and training is completed;

[0007] S200, the anchoring hole image is input into the trained structural plane pixel recognition model, the structural plane pixels of the anchoring hole image are recognized using the structural plane pixel recognition model, a thinning algorithm is used to reduce the number of structural plane pixels and maintain the skeleton of the original structural plane pixels, so as to obtain an anchoring hole image containing only structural plane skeleton pixels and other pixels;

[0008] S300, using the binomial value method to label the structural surface skeleton pixels and other pixels in the anchor hole image, and segmenting each column of structural surface skeleton pixels by judging whether each column of structural surface skeleton pixels in the anchor hole image is continuous to realize the segmentation of each column of structural surface skeleton pixels, so that the continuous structural surface skeleton pixels in each column are divided into a segment;

[0009] Specifically, the coordinates of each column of structural surface skeleton pixels in the anchor hole image are stored in a matrix, each column of structural surface skeleton pixels is judged one by one, if the row number corresponding to the column of structural surface skeleton pixels in the anchor hole image is continuous, the column of structural surface skeleton pixels is not split, if not continuous, the column of structural surface skeleton pixels is split at the discontinuous position, and each segment of the split structural surface skeleton pixels is stored in a different matrix, and the row number corresponding to the structural surface skeleton pixels in each matrix is continuous.

[0010] S400, judging whether two columns of structural surface skeleton pixels are adjacent column by column, if the two columns of structural surface skeleton pixels are adjacent, merging the two columns of structural surface skeleton pixels until the merged structural surface skeleton pixels are not adjacent to the next column of structural surface skeleton pixels, or the structural surface skeleton pixels intersect, completing the extraction of a structural surface skeleton pixel segment; extracting the coordinates of each structural surface skeleton pixel in all structural surface skeleton pixel segments and storing them in different matrices respectively, that is, the coordinates of the structural surface skeleton pixels of each structural surface skeleton pixel segment correspond to a matrix, which is recorded as a structural surface pixel segment coordinate matrix.

[0011] S500, fitting all structural surface skeleton pixel segment coordinates using a sine function model, the greater the fitting coefficient, the greater the probability that these structural surface skeleton pixel segments constitute a structural surface skeleton, and finally determining which structural surface pixel segment coordinate matrices can constitute a structural surface skeleton, and the obtained sine function model expression is the structural surface skeleton expression.

[0012] S600, outputting the structural surface parameters in the borehole according to the structural surface skeleton expression.

[0013] Further, a large number of anchor hole images are collected on site, a neural network is used to establish a structural surface pixel recognition model, a large number of anchor hole images are obtained by a borehole imager, the structural surface in the anchor hole is labeled with structural surface pixels to form a database, and then the neural network is trained to obtain the structural surface pixel recognition model.

[0014] Further, the binomial value method is used to label the structural surface skeleton pixels and other pixels in the anchor hole image, and each column of structural surface skeleton pixels is segmented by judging whether each column of structural surface skeleton pixels is continuous, so that the continuous structural surface skeleton pixels in each column are divided into a segment.

[0015] a. Structural surface skeleton pixels are represented by the value "0", and other pixels are represented by "1";

[0016] b. Obtain the position coordinates of the "0" value pixels, and store these pixel coordinates in a two-dimensional matrix R. The first column coordinates of the two-dimensional matrix R correspond to the row numbers of the "0" value pixel coordinates in the anchoring hole image, and the second column coordinates of the two-dimensional matrix R correspond to the column numbers of the "0" value pixel coordinates in the anchoring hole image;

[0017] c. If multiple "0" values in the two-dimensional matrix R are located in the same column of the anchoring hole image, resulting in multiple repeated column numbers of the anchoring hole image in the second column values of the matrix R, then the second column values of the two-dimensional matrix R are copied to the matrix R1. When there are repeated values in the matrix R1, only one value is retained. The matrix R1 stores the column numbers of the structure surface skeleton pixels in the anchoring hole image, thereby obtaining the number n of elements in the matrix R1;

[0018] d. Using the matrix R1 with the number n of elements, an empty cell B0 with n rows and 1 column is created, and then the structure surface skeleton pixel coordinates with the same second column values in the two-dimensional matrix R are copied to the cell B0, so that the cell B0 contains multiple matrices, each of which stores a column of structure surface skeleton pixel coordinates of the anchoring hole image;

[0019] e. According to whether the longitudinal coordinates of each column of structure surface skeleton pixels in the anchoring hole image are continuous, the matrices in the cell B0 are further split. Each column of continuous structure surface skeleton pixels is divided into matrices. If the i-th column has only one matrix of continuous structure surface skeleton pixels, the matrix is represented as B0{i}(1). If the i-th column is divided into m matrices of continuous structure surface skeleton pixels, these matrices are represented as B0{i}(1), B0{i}(2), …, B0{i}(m) respectively. Thus, the continuous structure surface skeleton pixel coordinate information in each column is stored in a separate matrix.

[0020] Further, the specific process of judging whether two columns of structure surface skeleton pixels are adjacent is to compare two matrices using formula (1):

[0021]

[0022] In the formula, x1 and x2 respectively represent the horizontal coordinates of the fissure pixel positions in the two matrices; y a and y b respectively represent the longitudinal coordinates of the fissure pixel positions in the two matrices; the horizontal coordinates of the stored coordinates in each matrix are the same, and the longitudinal coordinate values are different; K represents the minimum distance between the two matrices;

[0023] When the minimum distance between the two matrices is K, it is determined that the two matrices compared are adjacent: when the minimum distance between the two matrices is K, it is determined that the two matrices compared are not adjacent. ​If the structural plane skeleton pixel intersects, there must exist a matrix, which is less than or equal to the minimum distance between two or more matrices of the next column Or there are two or more matrices in the column, and the minimum distance between them and the same matrix of the next column is less than or equal to

[0024] Further, the coordinates of each structural plane skeleton pixel segment are fitted by a sine function model to finally determine the structural plane skeleton pixel, and the specific process is as follows:

[0025] a. Search for the structural plane pixel segment coordinate matrix containing the most structural plane skeleton pixels from the structural plane pixel segment coordinate matrix, and mark it as matrix C1;

[0026] b. Traverse all the structural plane skeleton pixel segment coordinate matrices. When the horizontal coordinates in the structural plane skeleton pixel segment coordinate matrix do not have repeated values with the horizontal coordinates in matrix C1, use the sine function model: Fit the structural plane skeleton pixel segment coordinate matrix with the coordinate points in matrix C1. In the sine function model, A represents the amplitude of the structural plane curve, The initial phase of the structural plane curve is represented by k, and P1 represents the horizontal resolution of the anchoring hole image.

[0027] c. Select the structural plane skeleton pixel segment coordinate matrix with the highest fitting coefficient and greater than 0.9, and add all the coordinates in it to matrix C1;

[0028] d. Repeat the traversal of all the structural plane skeleton pixel segment coordinate matrices, and store all the structural plane skeleton pixel coordinates that can be composed in C1. When the horizontal coordinates in the structural plane skeleton pixel segment coordinate matrix do not have repeated values with the horizontal coordinates in matrix C1, fit the structural plane skeleton pixel segment coordinate matrix with the coordinate points in matrix C1. Select the structural plane skeleton pixel segment coordinate matrix with the highest fitting coefficient and greater than 0.9, and add all the coordinates in it to matrix C1, until the number of structural plane skeleton pixel coordinates in matrix C1 no longer increases;

[0029] e. Statistics of the current matrix C1 corresponding to the structural plane skeleton pixel distribution in the anchoring hole image, when the matrix C1 corresponding to the structural plane skeleton pixel accounts for more than 80% of the total number of columns in the anchoring hole image, the structural plane skeleton pixel corresponding to the C1 matrix is regarded as a structural plane;

[0030] f. Since there may be multiple structural planes in the anchoring hole image, repeat steps a-e to search for other structural planes. Delete the structural plane skeleton pixel segment coordinate matrix corresponding to matrix C1 before repeating steps a-e each time, and reassign the value to matrix C1;

[0031] g. By steps a-f, it is judged whether there is still a structural surface skeleton pixel segment coordinate matrix, if not, it is ended, if yes, it is judged whether there is a coordinate in the structural surface pixel segment coordinate matrix located at the anchor hole boundary, if yes, it is judged whether there is a structural surface skeleton pixel segment coordinate matrix at the adjacent anchor hole image boundary and also has a coordinate located at the anchor hole boundary in the matrix, and the structural surface skeleton pixels corresponding to the structural surface skeleton pixel segment coordinate matrix at the adjacent two anchor hole boundaries are judged by steps b-e whether they can constitute a structural surface.

[0032] Further, according to the structural surface skeleton expression obtained by fitting, the structural surface parameters are output, including: structural surface inclination α, structural surface inclination angle β, structural surface position L, and the formula is as follows:

[0033]

[0034] In the formula, P2 represents the horizontal resolution and vertical resolution of the anchor hole image, D represents the anchor hole diameter, h represents the actual depth of the anchor hole, A represents the structural surface curve amplitude, φ represents the initial phase of the structural surface curve, and k represents the constant term of the structural surface curve. In the formula, P2 represents the horizontal resolution and vertical resolution of the anchor hole image, D represents the anchor hole diameter, h represents the actual depth of the anchor hole, A represents the structural surface curve amplitude, φ represents the initial phase of the structural surface curve, and k represents the constant term of the structural surface curve.

[0035] A computer device, comprising: a processor, a memory and a network interface;

[0036] The processor is connected with the memory and the network interface, wherein the network interface is used for providing network communication function, the memory is used for storing program code, and the processor is used for calling the program code to execute the method for quickly obtaining structural surface parameters in a borehole.

[0037] A computer readable storage medium, the computer readable storage medium stores a computer program, the computer program is suitable for being loaded by a processor and executing a method for quickly obtaining structural surface parameters in a borehole.

[0038] Compared with the prior art, the present application has the following advantages and beneficial effects:

[0039] 1. The present application proposes a new method for quickly extracting structural surface parameters, which considers various complex situations such as the influence of structural surface on rock layer interface, discontinuity of structural surface, intersection of structural surfaces and structural surface at the image boundary of anchor hole, and the recognition efficiency of the method is higher.

[0040] 2. The present application proposes a judgment method for two columns of structural surface skeleton pixels, realizes the extraction of continuous structural surface skeleton pixels in each column and the extraction of structural surface skeleton pixel segments, and lays a foundation for subsequent structural surface pixel determination.

[0041] 3、The application provides a new structural surface parameter rapid extraction method, which can accurately obtain structural surface pixels and further obtain structural surface parameters. BRIEF DESCRIPTION OF DRAWINGS

[0042] A structural surface parameter rapid acquisition method of the application is described below with reference to the drawings. In the drawings:

[0043] Figure 1 Fig. 1 is a flowchart of a structural surface parameter rapid acquisition method of the application.

[0044] Figure 2 Fig. 2 is a schematic diagram of a structural surface pixel determination process, wherein (a) is a schematic diagram of a structural surface pixel recognition result of an anchoring hole image, (b) is a schematic diagram of a structural surface skeleton extraction result, (c) is a schematic diagram of a structural surface skeleton pixel segment extraction result, and (d) is a schematic diagram of a structural surface skeleton pixel segment matching. DETAILED DESCRIPTION

[0045] The embodiments of the application are further described below with reference to the drawings:

[0046] As shown in Fig. 1, a structural surface parameter rapid acquisition method of a borehole comprises the following steps: Figure 1

[0047] S100, a large number of in-situ anchoring hole images are collected, a structural surface pixel recognition model is established by using a neural network, and training is completed.

[0048] A large number of anchoring hole images are collected in-situ by using a borehole imager to form a sample database, the sample database should include various rock layers such as sandstone, mudstone, sandy mudstone, and limestone. The structural surface pixels in the sample database are labeled, sample training is performed by using a neural network, and a structural surface pixel recognition model is obtained.

[0049] S200, the anchoring hole image is input into the trained structural surface pixel recognition model, the structural surface pixels of the anchoring hole image are recognized by using the structural surface pixel recognition model, a thinning algorithm is used to reduce the number of structural surface pixels and maintain the skeleton of the original structural surface pixels, and thus an anchoring hole image containing only structural surface skeleton pixels and other pixels is obtained.

[0050] The anchoring hole image is input into the trained structural surface pixel recognition model, and the structural surface pixels can be recognized. As shown in (a) of Fig. 2, the recognized structural surface pixels are shown in the schematic diagram, the black pixels in the figure are structural surface pixels, and two structural surfaces intersect with each other. The structural surface pixels are processed by using a thinning algorithm, the number of structural surface pixels can be reduced, but the skeleton of the original structural surface pixels is maintained. As shown in (b) of Fig. 2, the structural surface pixels are processed by using the thinning algorithm, and the structural surface pixels are shown in the schematic diagram. Figure 2 Figure 2 ​​(b) shows the structure surface pixel skeleton obtained by the thinning algorithm. The reduced number of structure surfaces is conducive to reducing the subsequent calculation amount and improving the structure surface recognition efficiency.

[0051] S300, using the binomial value method to label the structure surface skeleton pixels and other pixels in the image of the anchoring hole, segmenting each column of structure surface skeleton pixels by judging whether each column of structure surface skeleton pixels is continuous to realize the segmentation of each column of structure surface skeleton pixels, so as to divide the continuous structure surface skeleton pixels in each column into a segment.

[0052] Specifically, the coordinates of each column of structure surface skeleton pixels in the anchoring hole image are stored in a matrix, each column of structure surface skeleton pixels is judged one by one, if the corresponding row number of the column of structure surface skeleton pixels in the anchoring hole image is continuous, it is not split, if it is not continuous, it is split at the discontinuous place, and each segment of structure surface skeleton pixels after splitting is stored in a different matrix, the row number corresponding to the structure surface skeleton pixels in each matrix is continuous, and the specific steps include:

[0053] a. The structure surface skeleton pixels are represented by the numerical value "0", and the other pixels are represented by "1";

[0054] b. Obtain the position coordinates of the "0" value pixels, and store these pixel coordinates in a two-dimensional matrix R. The first column coordinates of the matrix R correspond to the row number of the "0" value pixel coordinates in the anchoring hole image, and the second column coordinates of the matrix R correspond to the column number of the "0" value pixel coordinates in the anchoring hole image. R is a two-dimensional matrix for storing coordinate values;

[0055] c. Because there are multiple "0" values in the same column of the anchoring hole image, the second column value of the matrix R will have multiple repeated column numbers of the anchoring hole image. Copy the second column value of the matrix R to the matrix R1. When there are repeated values in R1, only one value is retained, that is, the matrix stores the column number of the structure surface skeleton pixel in the anchoring hole image;

[0056] d. Obtain the number n of elements in the matrix R1, create an empty cell B0 with n rows and 1 column, then copy the structure surface skeleton pixel coordinates with the same second column value in the matrix R to the cell B0. At this time, the cell B0 contains multiple matrices, and each matrix stores the structure surface skeleton pixel coordinates of a column of the anchoring hole image;

[0057] e、According to whether the vertical coordinates of each column of structural surface skeleton pixels in the anchoring hole image are continuous, the matrix in the cell B0 is further split, and the structural surface skeleton pixels with continuous vertical coordinates in each column are divided into a matrix. If the i-th column has only one structural surface skeleton pixel matrix with continuous vertical coordinates, the matrix is denoted as B0{i}(1). If the i-th column is divided into m structural surface skeleton pixel matrices with continuous vertical coordinates, the matrices are denoted as B0{i}(1), B0{i}(2), …, B0{i}(m) respectively. Thus, the continuous structural surface skeleton pixel coordinate information in each column is stored in a matrix.

[0058] S400, whether two columns of structural surface skeleton pixels are adjacent is determined column by column. If two columns of structural surface skeleton pixels are adjacent, the two columns of structural surface skeleton pixels are merged until the merged structural surface skeleton pixels are not adjacent to the structural surface skeleton pixels in the next column, or the structural surface skeleton pixels intersect, and extraction of a structural surface skeleton pixel segment is completed. The coordinates of each structural surface skeleton pixel in all structural surface skeleton pixel segments are extracted and respectively stored in different matrices, that is, the structural surface skeleton pixel coordinates of each structural surface skeleton pixel segment correspond to a matrix, which is referred to as a structural surface pixel segment coordinate matrix.

[0059] Suppose that a structural surface skeleton pixel coordinate in a column is (x1, y1). If there is a structural surface skeleton pixel adjacent to (x1, y1) in the next column, there are three positions, which are (x1+1, y1), (x1+1, y1+1) and (x1+1, y1-1). The distances between the three positions and (x1, y1) are 1, 2 and 2 respectively. and Therefore, when two columns of structural surface pixels are adjacent, the minimum distance between the two columns of structural surface pixels should be less than or equal to

[0060] The specific process of determining whether two columns of structural surface skeleton pixels are adjacent column by column is to compare two matrices by using formula (1). When it is determined that the two columns of structural surface skeleton pixels are adjacent: when it is determined that the two columns of structural surface skeleton pixels are not adjacent: if the structural surface skeleton pixels intersect, there must be a matrix, the minimum distances between the matrix and two or more matrices in the next column are all less than or equal to or there are two or more matrices in the column, and the minimum distances between the two or more matrices and the same matrix in the next column are all less than or equal to

[0061]

[0062] In the formula, x1 and x2 respectively represent the horizontal coordinates corresponding to the positions of crack pixels in the two matrices; y a and y brespectively represent the vertical coordinates of the crack pixel positions in the two matrices; the horizontal coordinates of the stored coordinates in each matrix are the same, and the vertical coordinate values are different; K is the minimum distance between the two matrices.

[0063] The structural plane skeleton pixel segment extraction refers to that if a segment of structural plane skeleton pixels is continuous and does not intersect with other structural plane skeleton pixels, the coordinates of the segment of structural plane skeleton pixels are stored in a matrix. If a segment of continuous structural plane skeleton pixels intersects with other segments of structural plane skeleton pixels, the structural plane skeleton pixels are divided into multiple segments with the intersection points as the division points, and the coordinates of each segment of structural plane skeleton pixels exist in different matrices. The implementation method of the idea is to judge whether two columns of structural plane skeleton pixels are adjacent column by column, and if adjacent, to merge them into one segment of structural plane skeleton pixels. If the structural plane skeleton pixels intersect or are not adjacent, the extraction of the previous segment of structural plane skeleton pixels is completed, and the extraction of the next segment of structural plane skeleton pixels is started. Each segment of structural plane skeleton pixels is stored in a matrix. For example, Figure 2 As shown in (c) of FIG. 6, the two structural plane skeleton pixels are divided into six segments of structural plane skeleton pixels, which are respectively stored in matrices A1, A2, A3, A4, A5 and A6.

[0064] S500, the coordinates of each segment of structural plane skeleton pixels are fitted by using a sine function model. The greater the fitting coefficient is, the greater the probability that the structural plane skeleton pixels form a structural plane skeleton is. Finally, it is determined which coordinate matrix of the structural plane pixel segment corresponds to the coordinates that can form a structural plane skeleton. The expression of the sine function model obtained is the expression of the structural plane skeleton.

[0065] The specific process is as follows:

[0066] a. Using a sine function model: Search for the structural plane pixel segment coordinate matrix containing the most structural plane skeleton pixels from the structural plane pixel segment coordinate matrix, and mark it as matrix C1.

[0067] b. Traverse all the structural plane skeleton pixel segment coordinate matrices. When the horizontal coordinates in the structural plane skeleton pixel segment coordinate matrix do not have repeated values with the horizontal coordinates in matrix C1, use a sine function model: Fit the structural plane skeleton pixel segment coordinate matrix with the coordinate points in matrix C1. In the sine function model, A represents the amplitude of the structural plane curve, represents the initial phase of the structural plane curve, k represents the constant term of the structural plane curve, and P1 represents the horizontal resolution of the anchoring hole image.

[0068] c. Select the structural plane skeleton pixel segment coordinate matrix with the highest fitting coefficient and greater than 0.9, and add all the coordinates in the matrix to matrix C1.

[0069] ​d. In order to store all the structural plane skeleton pixel coordinates in matrix C1, repeat the traversal of all structural plane skeleton pixel segment coordinate matrices. When the horizontal coordinates in the structural plane skeleton pixel segment coordinate matrix are not repeated in matrix C1, fit the structural plane skeleton pixel segment coordinate matrix with the coordinate points in matrix C1. Select the structural plane skeleton pixel segment coordinate matrix with the highest fitting coefficient and greater than 0.9. Add all the coordinates in the matrix to matrix C1. Repeat the process until the number of structural plane skeleton pixel coordinates in matrix C1 no longer increases.

[0070] e. Count the number of structural plane skeleton pixels in matrix C1 that are distributed in the columns of the anchoring hole image. If the structural plane skeleton pixels in matrix C1 account for more than 80% of the total number of columns in the anchoring hole image, the structural plane skeleton pixels in matrix C1 are considered to be a structural plane.

[0071] Suppose there are 7 structural plane skeleton pixel segments in an anchoring hole image, and one structural plane has 4 structural plane skeleton pixel segments (x1, x2, x3, and x4). Through step a, select the structural plane skeleton pixel segment x1 with the most structural plane skeleton pixels from the 4 structural plane skeleton pixel segments. Through step b, fit the remaining 6 structural plane pixel segments with x1. Through step c, determine x2. Through step d, repeat the traversal of the first time to determine x3. Add x3 to matrix C1. Repeat the traversal of the second time to add x4 to matrix C1. When repeating the traversal again, the coordinates in matrix C1 no longer increase. At this time, x1, x2, x3, and x4 in matrix C1 are a structural plane.

[0072] f. There may be multiple structural planes in the anchoring hole image. Repeat steps a-e to search for other structural planes. Delete the structural plane skeleton pixel segment coordinate matrix corresponding to matrix C1 before repeating steps a-e each time.

[0073] g. Through steps a-f, determine whether there are still structural plane skeleton pixel segment coordinate matrices. If not, end the process. If there are, determine whether there are coordinates in the structural plane pixel segment coordinate matrix located at the boundary of the anchoring hole. If there are, determine whether there are structural plane skeleton pixel segment coordinate matrices at the boundaries of adjacent anchoring holes and whether the matrices also have coordinates located at the boundary of the anchoring hole. Use the above method to determine whether the structural plane skeleton pixel segment coordinate matrices at the boundaries of the two adjacent anchoring holes can form a structural plane.

[0074] According to the process, there are 4 matching schemes for the structural plane skeleton pixel segment coordinate matrix A1, as shown in Figure 2The matching of matrix A1 and A4, matrix A1 and A3, matrix A1 and A5, and matrix A1 and A6 is shown in (d). The matching scheme is fitted by using a sine function model, and it is confirmed that the fitting effect of matrix A1 and A4 and matrix A1 and A5 is good. The fracture matrix A1, A4 and A5 are fitted to determine all the pixels of the structure surface skeleton. The remaining matrix is matched and fitted by using a similar method to determine that A2, A3 and A6 belong to the second structure surface.

[0075] S600, outputting the structure surface parameters according to the structure surface skeleton expression.

[0076] After obtaining the structure surface skeleton expression, the structure surface dip angle, dip direction and position can be calculated by the following equation according to the sine function parameter value, so as to quickly output the structure surface parameters.

[0077]

[0078] In the formula, P2 represents the lateral resolution and longitudinal resolution of the anchoring hole image; D is the anchoring hole diameter; h is the actual depth of the anchoring hole; α and β are the structure surface dip direction and dip angle respectively; and L represents the structure surface position.

Claims

1. A method for quickly obtaining structural plane parameters in a borehole, characterized in that, The specific steps are as follows: S100, collect a large number of in-situ anchoring hole images, establish a structural plane pixel recognition model using a neural network and complete training; S200, input the anchoring hole image into the trained structural plane pixel recognition model, identify the structural plane pixels of the anchoring hole image using the structural plane pixel recognition model, reduce the number of structural plane pixels using a thinning algorithm and maintain the skeleton of the original structural plane pixels, thereby obtaining an anchoring hole image containing only structural plane skeleton pixels and other pixels; S300, label the structural plane skeleton pixels and other pixels of the anchoring hole image using a binomial value method, segment each column of structural plane skeleton pixels by judging whether the structural plane skeleton pixels in each column are continuous, thereby dividing the continuous structural plane skeleton pixels in each column into a segment; Specifically, the coordinates of the structural plane skeleton pixels in each column of the anchoring hole image are present in a matrix, each column of structural plane skeleton pixels is judged one by one, if the column of structural plane skeleton pixels is continuous in the corresponding row number of the anchoring hole image, it is not split, if it is not continuous, it is split at the discontinuous place, and each segment of structural plane skeleton pixels after splitting is present in a different matrix, and the structural plane skeleton pixels in each matrix are continuous in the corresponding row number; S400, judge whether two columns of structural plane skeleton pixels are adjacent column by column, if the two columns of structural plane skeleton pixels are adjacent, merge the two columns of structural plane skeleton pixels, until the merged structural plane skeleton pixels are not adjacent to the next column of structural plane skeleton pixels, or the structural plane skeleton pixels intersect, complete the extraction of a structural plane skeleton pixel segment; extract the coordinates of each structural plane skeleton pixel in all structural plane skeleton pixel segments and present them in different matrices respectively, that is, the coordinates of the structural plane skeleton pixels of each structural plane skeleton pixel segment correspond to a matrix, which is denoted as a structural plane pixel segment coordinate matrix; S500, fit all structural plane skeleton pixel segment coordinates using a sine function model, the larger the fitting coefficient, the greater the probability that these structural plane skeleton pixel segments constitute a structural plane skeleton, finally determine which structural plane pixel segment coordinate matrices can constitute a structural plane skeleton, and the obtained sine function model expression is the structural plane skeleton expression; S600, output the structural plane parameters in the borehole according to the structural plane skeleton expression.

2. The method according to claim 1, wherein a large number of images of anchor holes are collected on site, and a structural plane pixel recognition model is established by using a neural network. A large number of anchoring hole images are obtained by a borehole imager, the structural plane pixels in the anchoring hole are labeled to form a database, and then a neural network is trained to obtain a structural plane pixel recognition model.

3. The method according to claim 1, wherein The binomial value method is used to label the structural plane skeleton pixels and other pixels of the anchoring hole image, and each column of structural plane skeleton pixels is segmented by judging whether the structural plane skeleton pixels in each column are continuous, thereby dividing the continuous structural plane skeleton pixels in each column into a segment, and the specific process is as follows: a, the structural plane skeleton pixels are represented by the value "0", and the other pixels are represented by "1"; b. Obtain the position coordinates of the "0" value pixels, and store these pixel coordinates in a two-dimensional matrix R. The first column coordinates of the two-dimensional matrix R correspond to the row numbers of the "0" value pixels in the anchor hole image, and the second column coordinates of the two-dimensional matrix R correspond to the column numbers of the "0" value pixels in the anchor hole image; c. If there are multiple "0" values in the same column of the anchor hole image in the two-dimensional matrix R, resulting in multiple repeated column numbers of the anchor hole image in the second column of the matrix R, then copy the second column values of the two-dimensional matrix R to the matrix R1. When there are repeated values in the matrix R1, only one value is retained. The matrix R1 stores the column numbers of the anchor hole image where the structure surface skeleton pixels are located, thereby obtaining the number n of elements in the matrix R1; d. Using the matrix R1 with n elements, create an empty cell B0 with n rows and 1 column, and then copy the structure surface skeleton pixel coordinates with the same second column values in the two-dimensional matrix R to the cell B0, so that the cell B0 contains multiple matrices, each of which stores a column of structure surface skeleton pixel coordinates of the anchor hole image; e. According to whether the longitudinal coordinates of each column of structure surface skeleton pixels in the anchor hole image are continuous, the matrices in the cell B0 are further split. Each column of continuous structure surface skeleton pixels is divided into a matrix. If the i-th column has only one matrix of continuous structure surface skeleton pixels, then the matrix is represented as B0{i}(1). If the i-th column is divided into m matrices of continuous structure surface skeleton pixels, then these matrices are represented as B0{i}(1), B0{i}(2), …, B0{i}(m), respectively. Thus, the continuous structure surface skeleton pixel coordinate information in each column is stored in a separate matrix.

4. The method according to claim 1, wherein, The specific process of judging whether two columns of structure surface skeleton pixels are adjacent is to compare two matrices using formula (1): In the formula, x1 and x2 respectively represent the horizontal coordinates corresponding to the crack pixel positions in the two matrices; y a and y b respectively represent the vertical coordinates corresponding to the crack pixel positions in the two matrices; the horizontal coordinates of the stored coordinates in each matrix are the same, and the vertical coordinate values are different; K represents the minimum distance between the two matrices; If the minimum distance between two matrices is less than or equal to If the minimum distance between two matrices is less than or equal to If the minimum distance between two matrices is less than or equal to If the minimum distance between two matrices is less than or equal to 5. The method according to claim 1, wherein, Through the sine function model, the coordinates of each structure surface skeleton pixel segment are fitted, and finally the structure surface skeleton pixels are determined. The specific process is as follows: a. Search for the structure surface pixel segment coordinate matrix that contains the most structure surface skeleton pixels from the structure surface pixel segment coordinate matrix, and denote it as matrix C1; b、traverse all the structural surface skeleton pixel segment coordinate matrix, when the horizontal coordinate in the structural surface skeleton pixel segment coordinate matrix and the horizontal coordinate in the matrix C1 have no repeated values, use the sine function model: fit the structural surface skeleton pixel segment coordinate matrix with the coordinate points in the matrix C1; in the sine function model, A represents the amplitude of the structural surface curve, represents the initial phase of the structural surface curve, k represents the constant term of the structural surface curve, and P1 represents the lateral resolution of the anchoring hole image; c. Select the structure surface skeleton pixel segment coordinate matrix with the highest fitting coefficient and greater than 0.9, and add all the coordinates in it to matrix C1; d. Repeat the iteration of all structure surface skeleton pixel segment coordinate matrices, and store all the structure surface skeleton pixel coordinates that can be formed in C1. When the horizontal coordinates in the structure surface skeleton pixel segment coordinate matrix do not have repeated values in matrix C1, fit the structure surface skeleton pixel segment coordinate matrix with the coordinates in matrix C1, select the structure surface skeleton pixel segment coordinate matrix with the highest fitting coefficient and greater than 0.9, and add all the coordinates in it to matrix C1. This process continues until the number of structure surface skeleton pixel coordinates in matrix C1 no longer increases; e. Statistics the distribution of the structure surface skeleton pixels corresponding to the current matrix C1 in which columns of the anchor hole image. When the structure surface skeleton pixels corresponding to matrix C1 account for more than 80% of the total number of columns in the anchor hole image, the structure surface skeleton pixels corresponding to matrix C1 are considered as a structure surface. f、Since there may be multiple structural planes in the anchor hole image, repeat steps a-e to search for other structural planes, and delete the structural plane skeleton pixel segment coordinate matrix corresponding to matrix C1 before each repetition of steps a-e, and reassign values to matrix C1; g、Through steps a-f, determine whether there are still structural plane skeleton pixel segment coordinate matrices, if not, end; if yes, determine whether there are coordinates located at the anchor hole boundary in the structural plane skeleton pixel segment coordinate matrix, if yes, determine whether there are structural plane skeleton pixel segment coordinate matrices at the boundary of the adjacent anchor hole image and whether there are also coordinates located at the anchor hole boundary in the matrix, and determine whether the structural plane skeleton pixels corresponding to the structural plane skeleton pixel segment coordinate matrices at the boundaries of the two adjacent anchor holes can form a structural plane through steps b-e.

6. The method according to claim 5, wherein According to the structural plane skeleton expression obtained by fitting, output the structural plane parameters including the structural plane inclination α, the structural plane dip angle β and the structural plane position L, and the formula is as follows: In the formula, P2 represents the lateral resolution and the longitudinal resolution of the anchor hole image; D represents the anchor hole diameter; h represents the actual depth of the anchor hole, A represents the amplitude of the structural surface curve, and represents the initial phase of the structural surface curve, and k represents the constant term of the structural surface curve.

7. A computer device, comprising: It comprises: a processor, a memory and a network interface; The processor is connected with the memory and the network interface, wherein the network interface is used to provide network communication function, the memory is used to store program code, and the processor is used to call the program code to execute the borehole structural plane parameter fast acquisition method in any one of claims 1-6.

8. A computer-readable storage medium, characterized in that, The computer readable storage medium stores a computer program, and the computer program is suitable for being loaded and executed by the processor to execute the borehole structural plane parameter fast acquisition method in any one of claims 1-6.

Images