Pore network structure extraction method based on neighborhood search

Abstract

The invention discloses a pore network structure extraction method based on neighborhood search, and belongs to the field of porous medium model construction, and the method comprises the following steps: 1, obtaining a rock image, generating a binary model of a rock porous medium, removing isolated pores, and constructing an initial porous medium model; 2, randomly determining an initial search point, and generating adjacent elements and a neighborhood list of the initial search point; 3, determining the pore center of the porous medium model based on the initial search point; step 4, determining a maximum distance function gradient based on the adjacent element list so as to determine a central axis element set of the pore network structure; and step 5, combining the central axis element set to finally determine the pore center and the pore throat position of the porous medium model, and determining the whole pore network structure according to the pore center and the pore throat position. The method is high in calculation efficiency, and can achieve the efficient extraction of the pore network structure of the high-resolution rock image.

Description

Technical Field

[0001] The invention belongs to the field of porous medium model construction, and specifically relates to a pore network structure extraction method based on neighborhood search. Background Art

[0002] High-resolution rock images lead to low efficiency in extracting pore network structures, so there is an urgent need to develop efficient pore network structure extraction methods.

[0003] The development of imaging technology and machine learning methods has promoted the high resolution and high precision of underground rock images, which has brought challenges to the efficient extraction of pore network structures of porous media models. Therefore, it is particularly important to efficiently extract the pore network structure of high-precision porous media models. However, most of the current extraction methods process the porous media model as a whole, and the pixels or voxels of the model as a whole need to be calculated, which is inefficient and requires high computing resources. Therefore, how to achieve the extraction of pore network structure with less computing resources is an urgent problem to be solved. Summary of the invention

[0004] In order to solve the problem that the existing pore network extraction methods have low computational efficiency for the overall porous media model and are difficult to achieve efficient extraction of high-resolution models, the present invention proposes a pore network structure extraction method based on neighborhood search. This method targets the binary image of rock porous media, removes isolated pores and determines random initial search points, determines neighboring elements based on the search points and counts their neighborhood lists, and uses the ideas of gradient descent and neighborhood search to determine the pore center of the porous media, thereby determining the central axis position of the pore network, and finally determines the pore center and pore throat position of the porous media by combining the central axis elements.

[0005] The technical solution of the present invention is as follows: A pore network structure extraction method based on neighborhood search includes the following steps: Step 1: Obtain rock images, generate a binary model of rock porous media, remove isolated pores, and construct an initial porous media model; Step 2: Randomly determine the initial search point and generate the neighboring elements and neighborhood list of the initial search point; Step 3: Based on the initial search point, determine the pore center of the porous medium model; Step 4: Based on the list of neighboring elements, determine the maximum distance function gradient, thereby determining the central axis element set of the pore network structure; Step 5: Combine the central axis element set to finally determine the pore center and pore throat position of the porous medium model. The entire pore network structure can be determined based on the pore center and pore throat position.

[0006] Furthermore, the specific process of step 1 is as follows: Step 1.1, scan the rock image by core digital CT technology to obtain the initial raw file, denoise and binarize the raw file by Avizo software to generate a binary model of rock porous media; Step 1.2, determining the solid phase region and the pore phase region according to the binary model, and generating the target porous medium model; the specific generation process is: defining the position where the element value is 1 in the binary model as the solid phase, and defining the position where the element value is 0 as the pore phase, thereby determining the region of the solid phase and the pore phase in the target porous medium model, and then generating the target porous medium model; Step 1.3: Artificially set the open boundary conditions and closed boundary conditions of the target porous medium model to determine the connected pores and eliminate isolated pores: (1); In the formula, It is a connected pore element; It is an isolated pore element; is the pore element in the target porous media model; It is the set of elements after removing isolated pores; Step 1.4: Reconstruct the porous medium model based on the element set after eliminating isolated pores, and define the porous medium model as the initial porous medium model.

[0007] Furthermore, the specific process of step 2 is as follows: Step 2.1, randomly select an element in the initial porous medium model as the initial search point. The porous medium model is divided into a two-dimensional porous medium model and a three-dimensional porous medium model. The position coordinates of the initial search points of different dimensional models are defined as follows: (2); (3); In the formula, is the position coordinate of the initial search point in the two-dimensional porous media model, , are the initial search points in the two-dimensional porous media model. direction, Direction coordinates; is the position coordinate of the initial search point in the three-dimensional porous media model, , , are the initial search points in the three-dimensional porous media model. direction, direction, Direction coordinates; is a random function; Step 2.2, based on the initial search point, generate a neighboring element list and a solid neighboring domain list; Based on the initial search point, the neighboring element list and solid neighboring domain list of the two-dimensional porous media model are generated. The specific calculation formula is: (4); (5); (6); in, is the list of neighboring elements for the two-dimensional porous media model; Represents the set of solid phase elements within the judgment distance in the two-dimensional porous media model; is the position coordinate of the solid adjacent domain elements in the two-dimensional porous media model, , They are the solid adjacent domain elements in the two-dimensional porous media model. direction, Direction coordinates; , In the two-dimensional porous media model, direction, The judgment distance value in the direction; The collection of elements that represents the solid phase as a whole; is the solid adjacent domain element index, is the number of solid adjacent domain elements; The first The position coordinates of the adjacent domain elements of the solid, , They are respectively Solid adjacent domain elements direction, Direction coordinates; is the set of neighborhood elements of the initial search point for the solid phase in the two-dimensional porous media model; represents the cutoff radius; Based on the initial search point, the neighboring elements and solid neighboring domain lists of the three-dimensional porous media model are generated. The specific calculation formula is: (7); (8); (9); in, is the list of neighboring elements of the three-dimensional porous media model; It is the set of solid phase elements within the judgment distance in the three-dimensional porous media model; is the position coordinate of the solid adjacent domain elements in the three-dimensional porous media model, , , They are the solid adjacent domain elements in the two-dimensional porous media model. direction, direction, Direction coordinates; , , In the three-dimensional porous media model, direction, direction, The judgment distance value in the direction; is the set of neighborhood elements of the initial search point for the solid phase in the three-dimensional porous media model; The three-dimensional porous media model The position coordinates of the adjacent domain elements of the solid, , , They are respectively Solid adjacent domain elements direction, direction, Direction coordinates; The specific process of determining the solid adjacent domain list is as follows: first, determine the preliminary adjacent domain range according to the judgment distance values ​​in each direction; then, take the center point of the preliminary adjacent domain range as the center of the circular or spherical area, and As the radius, a circular or spherical region is obtained, and all solid phase elements within the region constitute the solid adjacent domain list.

[0008] Furthermore, the specific process of step 3 is as follows: Step 3.1: Based on the initial search point and the solid adjacent domain list information, the gradient descent method is used to determine the maximum gradient direction. The calculation formula is: (10); (11); (12); In the formula, Porous phase elements Solid phase elements The pore phase element is the initial search point, and the solid phase element is the element value in the solid adjacent domain list; Porous phase elements of Direction coordinates; Solid phase elements of Direction coordinates; Porous phase elements of Direction coordinates; Solid phase elements of Direction coordinates; Porous phase elements of Direction coordinates; Solid phase elements of Direction coordinates; is the gradient of the minimum distance function; is the set of pore phase elements; is the maximum distance function gradient; is the distance between the initial search point and the adjacent element; Step 3.2, determine the location of the initial pore center: (13); In the formula, The pore phase elements The pore space composed of The position coordinates of the elements, , The pore phase elements The pore space composed of Elements Direction coordinates, Direction coordinates; , Respectively element, The minimum distance function gradient of elements; is the position coordinate of the initial pore center, , The initial pore center Direction coordinates, Direction coordinates.

[0009] Furthermore, the specific process of step 4 is as follows: Step 4.1, starting from the initial search point, based on the list of neighboring elements, calculate the distance function gradient between the initial search point and each neighboring element according to formula (12), and obtain the maximum distance function gradient; Step 4.2, update the initial search point to the neighboring element corresponding to the maximum distance function gradient; Step 4.3: Based on the initial search point updated in step 4.2, steps 4.1 to 4.2 are repeated continuously, and the maximum distance function gradient of the adjacent elements is stored. The set of all updated initial search points generated during the updating process is the set of central axis elements starting from the initial search point.

[0010] Furthermore, the specific process of step 5 is as follows: Based on the central axis element set obtained in step 4, the local maximum and local minimum in the central axis element set are determined according to formula (14) and formula (15), respectively. The local minimum is the saddle point of the distance function value of the central axis element set, which is the pore throat; the local maximum is the peak value of the distance function value of the central axis element set, which is the pore center. The specific formula is: (14); (15); In the formula, , Respectively element, The maximum distance function gradient of elements; is the position coordinate of the pore center, , The pore center Direction coordinates, Direction coordinates; is the position coordinate of the pore throat, , The pore throat Direction coordinates, Direction coordinates.

[0011] The beneficial technical effects brought by the present invention are as follows: the method targets the binary model of rock porous media, removes isolated pores and determines random initial search points, determines neighboring elements based on the search points and counts their neighborhood lists, and determines the pore center of the porous media using the ideas of gradient descent and neighborhood search, which can determine the central axis position of the pore network. Combined with the central axis elements, the pore center and pore throat position of the porous media can be quickly determined, and the method has good application value. The present invention has high computational efficiency and can realize efficient extraction of the pore network structure of high-resolution rock images. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 It is a flow chart of the pore network structure extraction method based on neighborhood search of the present invention.

[0013] Figure 2 It is a schematic diagram of the framework of the pore network structure extraction method based on neighborhood search of the present invention.

[0014] Figure 3 It is a neighborhood determination graph of the pore network structure extraction method based on neighborhood search of the present invention.

[0015] Figure 4It is a model preprocessing diagram of the pore network structure extraction method based on neighborhood search adopted in an embodiment of the present invention.

[0016] Figure 5 It is a comparison diagram of the ideas of the method of the present invention and the conventional central axis method in the embodiment of the present invention, wherein (a) is the idea diagram of the method of the present invention, and (b) is the idea diagram of the conventional central axis method. DETAILED DESCRIPTION

[0017] The present invention is further described in detail below with reference to the accompanying drawings and specific embodiments: like Figure 1 and Figure 2 As shown, a pore network structure extraction method based on neighborhood search includes the following steps: Step 1: Obtain rock images, generate a binary model of rock porous media, remove isolated pores, and construct an initial porous media model; the specific process is as follows: Step 1.1, scan the rock image by core digital CT (computer tomography) technology to obtain the initial raw file, and use Avizo software to denoise and binarize the raw file to generate a binary model of rock porous media; Step 1.2, determine the solid phase region and the pore phase region according to the binary model, and generate the target porous medium model; the specific generation process is: define the position where the element value is 1 in the binary model as the solid phase, and define the position where the element value is 0 as the pore phase, so as to determine the region of the solid phase and the pore phase in the target porous medium model, and then generate the target porous medium model.

[0018] Step 1.3: Artificially set the open boundary conditions and closed boundary conditions of the target porous medium model to determine the connected pores and eliminate isolated pores: (1); In the formula, It is a connected pore element; It is an isolated pore element; is the pore element in the target porous media model; It is the set of elements after removing isolated pores.

[0019] Step 1.4: Reconstruct the porous medium model based on the element set after eliminating isolated pores, and define the porous medium model as the initial porous medium model.

[0020] Step 2: Randomly determine the initial search point and generate the neighboring elements and neighborhood list of the initial search point; the specific process is as follows: Step 2.1, randomly select an element in the initial porous medium model as the initial search point. In the present invention, the porous medium model can be a two-dimensional space model or a three-dimensional space model. The position coordinates of the initial search points of different dimensional models are defined as follows: (2); (3); In the formula, is the position coordinate of the initial search point in the two-dimensional porous media model, , are the initial search points in the two-dimensional porous media model. direction, Direction coordinates; is the position coordinate of the initial search point in the three-dimensional porous media model, , , are the initial search points in the three-dimensional porous media model. direction, direction, Direction coordinates; is a random function.

[0021] Step 2.2, based on the initial search point, generate a neighboring element list and a solid neighboring domain list; Based on the initial search point, the neighboring element list and solid neighboring domain list of the two-dimensional porous media model are generated. The specific calculation formula is: (4); (5); (6); in, is the list of neighboring elements for the two-dimensional porous media model; Represents the set of solid phase elements within the judgment distance in the two-dimensional porous media model; is the position coordinate of the solid adjacent domain elements in the two-dimensional porous media model, , They are the solid adjacent domain elements in the two-dimensional porous media model. direction, Direction coordinates; , In the two-dimensional porous media model, direction, The judgment distance value in the direction; The collection of elements that represents the solid phase as a whole; is the solid adjacent domain element index, is the number of solid adjacent domain elements; The first The position coordinates of the adjacent domain elements of the solid, , They are respectively Solid adjacent domain elements direction, Direction coordinates; is the set of neighborhood elements of the initial search point for the solid phase in the two-dimensional porous media model; Represents the cutoff radius.

[0022] Based on the initial search point, the neighboring elements and solid neighboring domain lists of the three-dimensional porous media model are generated. The specific calculation formula is: (7); (8); (9); in, is the list of neighboring elements of the three-dimensional porous media model; It is the set of solid phase elements within the judgment distance in the three-dimensional porous media model; is the position coordinate of the solid adjacent domain elements in the three-dimensional porous media model, , , They are the solid adjacent domain elements in the two-dimensional porous media model. direction, direction, Direction coordinates; , , In the three-dimensional porous media model, direction, direction, The judgment distance value in the direction; is the set of neighborhood elements of the initial search point for the solid phase in the three-dimensional porous media model; The three-dimensional porous media model The position coordinates of the adjacent domain elements of the solid, , , They are respectively Solid adjacent domain elements direction, direction, Direction coordinates; The specific process of determining the solid adjacent domain list is as follows: first, determine the preliminary adjacent domain range according to the judgment distance values ​​in each direction; then, take the center point of the preliminary adjacent domain range as the center of the circular or spherical area, and As the radius, a circular or spherical region is obtained, and all solid phase elements within the region constitute the solid adjacent domain list.

[0023] Figure 3 This is a specific process for determining the solid adjacent domain list in the two-dimensional porous medium model of the present invention. The left side of the arrow is the initial adjacent domain range determined by judging the distance value. , They are direction, The judgment distance value in the direction, represents the solid phase element set within the judgment distance; the figure on the right side of the arrow is the preliminary neighborhood range determined by judging the distance of the neighborhood. Based on the preliminary neighborhood range, the center point of the preliminary neighborhood range is taken as the center of the circular area. As the radius, a circular region is obtained, and all solid phase elements within the region constitute the solid adjacent domain list. is the set of neighborhood elements of the initial search point for the solid phase.

[0024] Step 3: Based on the initial search point, determine the pore center of the porous medium model; Step 3.1: Based on the initial search point and the solid adjacent domain list information, the gradient descent method is used to determine the maximum gradient direction. The calculation formula is: (10); (11); (12); In the formula, Porous phase elements Solid phase elements The pore phase element is the initial search point, and the solid phase element is the element value in the solid adjacent domain list; Porous phase elements of Direction coordinates; Solid phase elements of Direction coordinates; Porous phase elements of Direction coordinates; Solid phase elements of Direction coordinates; Porous phase elements of Direction coordinates; Solid phase elements of Direction coordinates; is the gradient of the minimum distance function; is the set of pore phase elements; is the maximum distance function gradient; is the distance between the initial search point and the adjacent element.

[0025] Step 3.2, determine the location of the initial pore center: 13); In the formula, The pore phase elements The pore space composed of The position coordinates of the elements, , The pore phase elements The pore space composed of Elements Direction coordinates, Direction coordinates; , Respectively element, The minimum distance function gradient of elements; is the position coordinate of the initial pore center, , The initial pore center Direction coordinates, Direction coordinates.

[0026] Step 4: Based on the list of neighboring elements, determine the maximum distance function gradient, thereby determining the central axis element set of the pore network structure; the specific process is: Step 4.1, starting from the initial search point, based on the list of neighboring elements, calculate the distance function gradient between the initial search point and each neighboring element according to formula (12), and obtain the maximum distance function gradient; Step 4.2, update the initial search point to the neighboring element corresponding to the maximum distance function gradient; Step 4.3: Based on the initial search point updated in step 4.2, steps 4.1 to 4.2 are repeated continuously, and the maximum distance function gradient of the adjacent elements is stored. The set of all updated initial search points generated during the updating process is the set of central axis elements starting from the initial search point.

[0027] Step 5. Combine the central axis element set to finally determine the pore center and pore throat position of the porous media model. The entire pore network structure can be determined based on the pore center and pore throat position. Based on the central axis element set obtained in step 4, the local maximum and local minimum in the central axis element set are determined according to formula (14) and formula (15) respectively. The local minimum is the saddle point of the distance function value of the central axis element set, which is the pore throat; the local maximum is the peak value of the distance function value of the central axis element set, which is the pore center. The specific formula is: 14); (15); In the formula, , Respectively element, The maximum distance function gradient of elements; is the position coordinate of the pore center, , The pore center Direction coordinates, Direction coordinates; is the position coordinate of the pore throat, , The pore throat Direction coordinates, Direction coordinates.

[0028] In order to demonstrate the feasibility and superiority of the present invention, the following specific examples are given.

[0029] The present invention uses a two-dimensional porous medium model as an example for testing, and the model size is 20 pixels × 20 pixels. Figure 4 This is a model preprocessing diagram of the pore network structure extraction method based on neighborhood search in the embodiment of the present invention. The porous medium model consists of a solid phase (gray squares) and a pore phase (white squares), with closed boundary conditions at the top and bottom and open boundary conditions at the left and right as inflow and outflow channels to determine the connected pore space. In addition to the connected pore phase elements, other isolated pore phase elements are changed to solid phase elements by changing their binary model values ​​from 0 to 1. Traverse the solid phase elements, if there are pore phase elements in the neighboring elements of this solid phase element, they are retained as solid phase boundary elements, otherwise they are elements within the solid phase and are eliminated (the binary model value is changed from 1 to 2). In this embodiment, the number of neighboring pixels for each pixel is set to 8, the neighborhood cutoff radius is 5 pixels, the search step is 1 pixel, and the judgment range of the pore center and the pore throat is set to 3 pixels.

[0030] Figure 51 is a diagram comparing the ideas of the method of the present invention and the conventional central axis method in an embodiment of the present invention, (a) is a diagram showing the idea of ​​the method of the present invention, and (b) is a diagram showing the idea of ​​the conventional central axis method. Figure 5 The numbers 1-4 represent the distance function gradient of the pore phase element from the solid phase. It can be seen that the conventional central axis method needs to calculate the distance function gradient of each pore phase element to obtain the position of the central axis, while the present invention only needs to calculate the neighboring pixels of the search point to obtain the central axis.

[0031] Of course, the above description is not a limitation of the present invention, and the present invention is not limited to the above examples. Changes, modifications, additions or substitutions made by technicians in this technical field within the essential scope of the present invention should also fall within the protection scope of the present invention.

Claims

1. A pore network structure extraction method based on neighborhood search, characterized in that: The steps include: Step 1: Obtain rock images, generate a binary model of rock porous media, remove isolated pores, and construct an initial porous media model; Step 2: Randomly determine the initial search point and generate the neighboring elements and neighborhood list of the initial search point; Step 3: Based on the initial search point, determine the pore center of the porous medium model; Step 4: Based on the list of neighboring elements, determine the maximum distance function gradient, thereby determining the central axis element set of the pore network structure; Step 5: Combine the central axis element set to finally determine the pore center and pore throat position of the porous medium model. The entire pore network structure can be determined based on the pore center and pore throat position.

2. The pore network structure extraction method based on neighborhood search according to claim 1 is characterized in that: The specific process of step 1 is as follows: Step 1.1, scan the rock image by core digital CT technology to obtain the initial raw file, denoise and binarize the raw file by Avizo software to generate a binary model of rock porous media; Step 1.2, determining the solid phase region and the pore phase region according to the binary model, and generating the target porous medium model; the specific generation process is: defining the position where the element value is 1 in the binary model as the solid phase, and defining the position where the element value is 0 as the pore phase, thereby determining the region of the solid phase and the pore phase in the target porous medium model, and then generating the target porous medium model; Step 1.3: Artificially set the open boundary conditions and closed boundary conditions of the target porous medium model to determine the connected pores and eliminate isolated pores: (1)； In the formula, It is a connected pore element; It is an isolated pore element; is the pore element in the target porous media model; It is the set of elements after removing isolated pores; Step 1.4: Reconstruct the porous medium model based on the element set after eliminating isolated pores, and define the porous medium model as the initial porous medium model.

3. The pore network structure extraction method based on neighborhood search according to claim 2 is characterized in that: The specific process of step 2 is: Step 2.1, randomly select an element in the initial porous medium model as the initial search point. The porous medium model is divided into a two-dimensional porous medium model and a three-dimensional porous medium model. The position coordinates of the initial search points of different dimensional models are defined as follows: (2)； (3)； In the formula, is the position coordinate of the initial search point in the two-dimensional porous media model, , are the initial search points in the two-dimensional porous media model. direction, Direction coordinates; is the position coordinate of the initial search point in the three-dimensional porous media model, , , are the initial search points in the three-dimensional porous media model. direction, direction, Direction coordinates; is a random function; Step 2.2, based on the initial search point, generate a neighboring element list and a solid neighboring domain list; Based on the initial search point, the neighboring element list and solid neighboring domain list of the two-dimensional porous media model are generated. The specific calculation formula is: (4)； (5)； (6)； in, is the list of neighboring elements for the two-dimensional porous media model; Represents the set of solid phase elements within the judgment distance in the two-dimensional porous media model; is the position coordinate of the solid adjacent domain elements in the two-dimensional porous media model, , They are the solid adjacent domain elements in the two-dimensional porous media model. direction, Direction coordinates; , In the two-dimensional porous media model, direction, The judgment distance value in the direction; The collection of elements that represents the solid phase as a whole; is the solid adjacent domain element index, is the number of solid adjacent domain elements; The first The position coordinates of the solid adjacent domain elements, , They are respectively Solid adjacent domain elements direction, Direction coordinates; is the set of neighborhood elements of the initial search point for the solid phase in the two-dimensional porous media model; represents the cutoff radius; Based on the initial search point, the neighboring elements and solid neighboring domain lists of the three-dimensional porous media model are generated. The specific calculation formula is: (7)； (8)； (9)； in, is the list of neighboring elements of the three-dimensional porous media model; It is the set of solid phase elements within the judgment distance in the three-dimensional porous media model; is the position coordinate of the solid adjacent domain elements in the three-dimensional porous media model, , , They are the solid adjacent domain elements in the two-dimensional porous media model. direction, direction, Direction coordinates; , , In the three-dimensional porous media model, direction, direction, The judgment distance value in the direction; is the set of neighborhood elements of the initial search point for the solid phase in the three-dimensional porous media model; The three-dimensional porous media model The position coordinates of the solid adjacent domain elements, , , They are respectively Solid adjacent domain elements direction, direction, Direction coordinates; The specific process of determining the solid adjacent domain list is as follows: first, determine the preliminary adjacent domain range according to the judgment distance values ​​in each direction; then, take the center point of the preliminary adjacent domain range as the center of the circular or spherical area, and As the radius, a circular or spherical region is obtained, and all solid phase elements within the region constitute the solid adjacent domain list.

4. The pore network structure extraction method based on neighborhood search according to claim 3 is characterized in that: The specific process of step 3 is as follows: Step 3.1: Based on the initial search point and the solid adjacent domain list information, the gradient descent method is used to determine the maximum gradient direction. The calculation formula is: (10)； (11)； (12)； In the formula, Porous phase elements Solid phase elements The pore phase element is the initial search point, and the solid phase element is the element value in the solid adjacent domain list; Porous phase elements of Direction coordinates; Solid phase elements of Direction coordinates; Porous phase elements of Direction coordinates; Solid phase elements of Direction coordinates; Porous phase elements of Direction coordinates; Solid phase elements of Direction coordinates; is the gradient of the minimum distance function; is the set of pore phase elements; is the maximum distance function gradient; is the distance between the initial search point and the adjacent element; Step 3.2, determine the location of the initial pore center: (13)； In the formula, The pore phase elements The pore space composed of The position coordinates of the elements, , The pore phase elements The pore space composed of Elements Direction coordinates, Direction coordinates; , Respectively element, The minimum distance function gradient of elements; is the position coordinate of the initial pore center, , The initial pore center Direction coordinates, Direction coordinates.

5. The method for extracting pore network structure based on neighborhood search according to claim 4, characterized in that: The specific process of step 4 is as follows: Step 4.1, starting from the initial search point, based on the list of neighboring elements, calculate the distance function gradient between the initial search point and each neighboring element according to formula (12), and obtain the maximum distance function gradient; Step 4.2, update the initial search point to the neighboring element corresponding to the maximum distance function gradient; Step 4.3: Based on the initial search point updated in step 4.2, steps 4.1 to 4.2 are repeated continuously, and the maximum distance function gradient of the adjacent elements is stored. The set of all updated initial search points generated during the updating process is the set of central axis elements starting from the initial search point.

6. The method for extracting pore network structure based on neighborhood search according to claim 5, characterized in that: The specific process of step 5 is as follows: Based on the central axis element set obtained in step 4, the local maximum and local minimum in the central axis element set are determined according to formula (14) and formula (15), respectively. The local minimum is the saddle point of the distance function value of the central axis element set, which is the pore throat; the local maximum is the peak value of the distance function value of the central axis element set, which is the pore center. The specific formula is: (14)； (15)； In the formula, , Respectively element, The maximum distance function gradient of elements; is the position coordinate of the pore center, , The pore center Direction coordinates, Direction coordinates; is the position coordinate of the pore throat, , The pore throat Direction coordinates, Direction coordinates.

Images