Method for generating local anisotropic search ellipsoid based on 3D-SOBEL

Abstract

The present invention relates to a method for generating a local anisotropic search ellipsoid based on 3D-SOBEL, including: S1. Obtaining the coordinate nodes and topological information of an ore body model; S2. Constructing an ore body grid model based on the coordinate nodes and topological information, performing an indicatrization process on the ore body grid model, calculating the gradient at each unit position in the ore body grid model, and obtaining the boundary information of the ore body grid model; S3. According to the boundary information, obtaining the ellipsoid shape at the boundary of the ore body grid model, calculating the ellipsoid shape parameters at the boundary of the ore body grid model, and generating a search ellipsoid model inside the ore body. The present invention determines the internal ellipsoid direction based on the ellipsoid direction at the boundary, and the calculation is relatively simple without a variogram or other complex calculation processes.

Description

Technical Field

[0001] The present invention relates to the technical field of geostatistical mineral resource reserve estimation, and particularly to a method for generating a local anisotropic search ellipsoid based on 3D-SOBEL. Background Art

[0002] Ore grade estimation requires selecting a reasonable estimation method and parameters by considering the geological characteristics of the ore body and the spatial distribution of exploration data. Among them, the parameter setting of the search ellipsoid will directly affect the spatial distribution and quantity of the samples to be estimated, and further affect the accuracy of the estimation result.

[0003] The traditional search ellipsoid setting mainly includes two methods: (1) relying on expert experience; (2) judging according to spatial variability. The first method is greatly affected by human factors; the second method requires calculating the variogram, and there is a certain degree of difficulty in the operation process. In addition, when the ore body has a complex spatial shape or complex variability, different search ellipsoids need to be set at different positions. Currently, the commonly used methods are difficult to meet this actual demand. Summary of the Invention

[0004] The purpose of the present invention is to provide a method for generating a local anisotropic search ellipsoid with low human factor influence and capable of setting different search ellipsoids at different positions to solve the defects existing in the above-mentioned prior art.

[0005] To achieve the above purpose, the present invention provides the following scheme:

[0006] A method for generating a local anisotropic search ellipsoid based on 3D-SOBEL, comprising:

[0007] S1. Obtain the coordinate nodes and topological information of the ore body model;

[0008] S2. Based on the coordinate nodes and topological information, construct an ore body grid model, perform an indicator processing on the ore body grid model, calculate the gradient of each unit position in the ore body grid model, and obtain the boundary information of the ore body grid model;

[0009] S3. According to the boundary information, obtain the ellipsoid shape at the boundary of the ore body grid model, calculate the ellipsoid shape parameters at the boundary of the ore body grid model, and generate a search ellipsoid model inside the ore body.

[0010] Preferably, the coordinate nodes include the coordinate positions of the points in the ore body model, and the topological information includes: obtaining the position points of the ore body model, and based on the position points, obtaining the position surfaces of the ore body model.

[0011] Preferably, constructing the ore body grid model includes:

[0012] Based on the coordinate nodes and topological information, place the ore body model in a cuboid, and divide the cuboid into a number of cuboid units with the same size. Based on the length, width, and height of the cuboid units, select the cuboid units to obtain the ore body grid model;

[0013] Among them, the cuboid completely wraps the ore body model.

[0014] Preferably, the process of indicating the grid model includes:

[0015] Based on the center points of each cuboid unit in the ore body grid model, use the ray method or the spatial partition binary tree method to distinguish the grid units inside the ore body model and the grid units outside the ore body model, and obtain the indicated grid model.

[0016] Preferably, obtaining the boundary information of the ore body grid model includes:

[0017] Use the 3D-SOBEL algorithm to calculate the ore body grid model to obtain a direction template, and perform convolution calculation on the direction template and the indicated grid model to obtain the boundary information of the ore body grid model.

[0018] Preferably, the method of the convolution calculation is:

[0019] Gx(x,y,z) = [f(x + 1,y - 1,z + 1)+2f(x + 1,y,z + 1)+f(x + 1,y + 1,z + 1)] - [f(x - 1,y - 1,z + 1)+2f(x - 1,y,z + 1)+f(x - 1,y + 1,z + 1)]+…

[0020] Gy(x,y,z) = [f(x - 1,y - 1,z + 1)+2f(x,y - 1,z + 1)+f(x + 1,y - 1,z + 1)] - [f(x - 1,y + 1,z + 1)+2f(x,y + 1,z + 1)+f(x + 1,y + 1,z + 1)]+…

[0021] Gz(x,y,z) = [f(x - 1,y - 1,z + 1)+2f(x - 1,y,z + 1)+f(x + 1,y + 1,z + 1)] - [f(x - 1,y - 1,z - 1)+2f(x - 1,y,z - 1)+f(x - 1,y + 1,z - 1)]+…

[0022] Among them, f(x,y,z) is the indication value of the cuboid unit, and (x,y,z) is the center point coordinate.

[0023] Preferably, calculating the ellipsoid shape at the boundary of the grid model includes:

[0024] According to the boundary information, obtain the grid model at the boundary, traverse the grid model at the boundary through a local window to obtain a gradient matrix, and based on the gradient matrix, obtain the ellipsoid shape parameters at the boundary of the grid model.

[0025] Preferably, the method for traversing the grid model at the boundary based on a local window is:

[0026]

[0027] where n is the number of gradient data in the local window, are the gradients in the x, y, and z directions of the i-th data respectively.

[0028] Preferably, generating the search ellipsoid model inside the ore body includes:

[0029] Based on the eigenvalues and eigenvectors of the gradient matrix, obtain the search ellipsoid shape parameters at the boundary;

[0030] Based on the search ellipsoid shape parameters at the boundary, use spatial interpolation to predict the search ellipsoid shape parameters inside the ore body to obtain the search ellipsoid model parameters inside the ore body, and generate the search ellipsoid model inside the ore body;

[0031] where the search ellipsoid shape parameters at the boundary include: the azimuths of the three main axes of the search ellipsoid model at the boundary, and the ratio of the lengths of the three main axes of the search ellipsoid model at the boundary.

[0032] The beneficial effects of the present invention are:

[0033] The search ellipsoid generation method proposed by the present invention has little influence of human factors, and the direction of the ellipsoid inside is determined by the direction of the ellipsoid at the boundary. The calculation is relatively simple, without a variogram or other complex calculation processes. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0035] Figure 1 It is a flowchart of a local search ellipsoid generation method based on 3D - SOBEL according to an embodiment of the present invention;

[0036] Figure 2 It is a schematic diagram of the 3D - SOBEL algorithm template according to an embodiment of the present invention;

[0037] Figure 3 Schematic diagram of the ore body solid model in the embodiment of the present invention;

[0038] Figure 4 Schematic diagram of the grid filling effect in the embodiment of the present invention;

[0039] Figure 5 Schematic diagram of the indication processing effect in the embodiment of the present invention;

[0040] Figure 6 Schematic diagram of the boundary ellipsoid effect in the embodiment of the present invention;

[0041] Figure 7 Schematic diagram of the effect after internal interpolation in the embodiment of the present invention; Detailed implementation manners

[0042] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0043] To make the above objects, features, and advantages of the present invention more obvious and understandable, the present invention will be further described in detail below in conjunction with the accompanying drawings and specific implementation manners.

[0044] According to Figure 1 as shown, the method for generating a local anisotropic search ellipsoid based on 3D - SOBEL includes:

[0045] S1. Obtain the coordinate nodes and topological information of the ore body model;

[0046] S2. Based on the coordinate nodes and topological information, construct an ore body grid model, perform indication processing on the ore body grid model, calculate the gradient at each unit position in the ore body grid model, and obtain the boundary information of the ore body grid model;

[0047] S3. According to the boundary information, obtain the ellipsoid shape at the boundary of the ore body grid model, calculate the ellipsoid shape parameters at the boundary of the ore body grid model, and generate a search ellipsoid model inside the ore body.

[0048] According to Figure 3 as shown, obtain the coordinate nodes and topological information of the ore body solid model:

[0049] The storage structure of the ore body solid model contains a lot of information, such as geometric data (point, line, and surface coordinates), attribute data, and texture material data. The purpose of this step is to display the shape and position of the ore body solid model. Only the coordinate nodes and topological information need to be obtained. Among them, obtaining the coordinate nodes and topological information of the ore body solid model: obtaining the coordinate nodes and topological information to determine the shape, size, and position of the ore body. Among them, the coordinate nodes include the coordinate positions of the points in the ore body model, and the topological information includes: obtaining the position points of the ore body model, and based on the position points, obtaining the position surfaces of the ore body model.

[0050] According to Figure 4 shown, perform grid filling on the ore body model:

[0051] First, place the ore body model in a cuboid of appropriate size, which completely encloses the ore body model. Then divide the cuboid into small cuboids of equal size. It is necessary to select small cuboids of appropriate and equal size, which are the grid models. If the small cuboids are too large, they cannot represent the ore body model well. If the small cuboids are too small, the calculation amount will increase.

[0052] According to Figure 5 shown, perform indication processing on the grid model:

[0053] Assign the grid model units inside and outside the ore body model the values of 0 and 1 respectively. It can be implemented using the ray method and the BSP (Binary Space Partitioning) method. The ray method is often used to judge the position relationship between a point and a surface in a two-dimensional space. Its main idea is: draw a ray in any direction starting from the target point, and calculate the number of intersection points between the ray and the contour line of the surface. If the number is even, it means outside the surface; if it is odd, it means inside the surface. When extending it to judge the position relationship between a point and a volume in a three-dimensional space, only need to change calculating the number of intersection points between the ray and the contour line of the surface to calculating the number of intersection points between the ray and the boundary surface of the volume. Similarly, an even number means the point is outside the volume, and an odd number means the point is inside the volume. The BSP (Binary Space Partitioning) method is an algorithm that can effectively judge the position relationship between a three-dimensional model and a polygon or polyhedron. It forms a tree with all the divided spaces. Each space divides the space it is in into two parts, the front and the back, and these two spaces are further divided into smaller spaces by other spaces until the end. Each of its nodes represents a divided space. The left subtree of the node represents the positive side of the divided space, the right subtree represents the negative side, and the leaf nodes represent the convex regions obtained by the division.

[0054] When assigning values to the grid model, an indication value of 1 means that the small cuboid is inside the ore body model, and an indication value of 0 means it is outside. When using the above method to assign values to the grid model, the center points of each small cuboid are used for calculation and judgment.

[0055] According to Figure 2As described above, the gradient of each unit position in the indication network model is calculated using the 3D - SOBEL algorithm:

[0056] The SOBEL algorithm is a common edge detection algorithm in image processing, and the 3D - SOBEL algorithm is its application in a three - dimensional scene. This algorithm uses templates in the horizontal and vertical directions to convolve with the corresponding image data to perform weighted calculations on discrete data. For a three - dimensional scene, in this embodiment, templates in the x, y, and z directions (as Figure 2 shown) are used for calculation to obtain the boundary information of the ore body model. The formula is as follows:

[0057] Gx(x,y,z) = [f(x + 1,y - 1,z + 1)+2f(x + 1,y,z + 1)+f(x + 1,y + 1,z + 1)] - [f(x - 1,y - 1,z + 1)+2f(x - 1,y,z + 1)+f(x - 1,y + 1,z + 1)]+…

[0058] Gy(x,y,z) = [f(x - 1,y - 1,z + 1)+2f(x,y - 1,z + 1)+f(x + 1,y - 1,z + 1)] - [f(x - 1,y + 1,z + 1)+2f(x,y + 1,z + 1)+f(x + 1,y + 1,z + 1)]+…

[0059] Gz(x,y,z) = [f(x - 1,y - 1,z + 1)+2f(x - 1,y,z + 1)+f(x + 1,y + 1,z + 1)] - [f(x - 1,y - 1,z - 1)+2f(x - 1,y,z - 1)+f(x - 1,y + 1,z - 1)]+…

[0060] Where f(x,y,z) is the indication value of the cuboid unit, and (x,y,z) is the center point coordinate.

[0061] Calculating the ellipsoidal shape at the boundary of the ore body model

[0062] According to Figure 6 shown, use a local window to traverse the indication network model (the grid model at the boundary), and calculate the search ellipsoidal shape of each unit grid at the boundary position of the ore body model through the following method: (1) Assume that the number of gradient data in the current local window is n, and generate the gradient matrix G in the local window according to the following formula lw ; (2) Calculate the eigenvalues and eigenvectors of the matrix G lw , and record the corresponding results as: λ1, λ2, λ3 (λ1 < λ2 < λ3) and v1, v2, v3. Then the orientations of the three main axes of the search ellipsoid at the corresponding position are (v1, v2, v3), and the ratio of the lengths of the three axes is

[0063]

[0064] Generate search ellipsoids at other positions inside the ore body through spatial interpolation.

[0065] The parameters of the search ellipsoids near the boundary of the ore body model are determined. For the cuboids inside the model, the direction and size of the search ellipsoids are taken as the attributes to be predicted. By the method of spatial interpolation, the parameters of the search ellipsoids at other positions inside the ore body are predicted. According to Figure 7 As shown, a search ellipsoid model inside the ore body is generated, which is all local search ellipsoid models with the small cuboids inside the ore body as the center points.

[0066] The embodiments described above are only descriptions of the preferred embodiments of the present invention, and do not limit the scope of the present invention. Without departing from the design spirit of the present invention, various deformations and improvements made by those of ordinary skill in the art to the technical solutions of the present invention shall fall within the protection scope determined by the claims of the present invention.

Claims

1. A method for generating a local anisotropic search ellipsoid based on 3D-SOBEL, characterized in that Including: S1. Obtain the coordinate nodes and topological information of the ore body model; S2. Based on the coordinate nodes and topological information, construct an ore body grid model, perform an indication process on the ore body grid model, calculate the gradient at each unit position in the ore body grid model, and obtain the boundary information of the ore body grid model; S3. According to the boundary information, obtain the ellipsoid shape at the boundary of the ore body grid model, calculate the ellipsoid shape parameters at the boundary of the ore body grid model, and generate a search ellipsoid model inside the ore body: According to the boundary information, obtain the grid model at the boundary, traverse the grid model at the boundary through a local window to obtain a gradient matrix, and based on the eigenvalues and eigenvectors of the gradient matrix, obtain the ellipsoid shape parameters; The method for traversing the grid model at the boundary based on a local window is: where n is the number of gradient data in the local window, are the gradients of the i-th data in the x, y, and z directions respectively; Based on the search ellipsoid shape parameters, use spatial interpolation to predict the search ellipsoid shape parameters inside the ore body, obtain the search ellipsoid model parameters inside the ore body, and generate a search ellipsoid model inside the ore body; Among them, the search ellipsoid shape parameters at the boundary include: the azimuths of the three main axes of the search ellipsoid model at the boundary, and the ratio of the lengths of the three main axes of the search ellipsoid model at the boundary.

2. The method for generating a local anisotropic search ellipsoid based on 3D-SOBEL according to claim 1, wherein The coordinate nodes include the coordinate positions of the points in the ore body model, and the topological information includes: obtaining the position points of the ore body model, and based on the position points, obtaining the position planes of the ore body model.

3. The method for generating a local anisotropic search ellipsoid based on 3D-SOBEL according to claim 1, characterized in that Constructing the ore body grid model includes: Based on the coordinate nodes and topological information, place the ore body model in a cuboid, and divide the cuboid into several cuboid units with the same size. Based on the length, width, and height of the cuboid units, select the cuboid units to obtain the ore body grid model; Among them, the cuboid completely wraps the ore body model.

4. The method for generating a local anisotropy search ellipsoid based on 3D-SOBEL according to claim 3, wherein Performing an indication process on the grid model includes: Based on the center points of each cuboid unit in the ore body grid model, use the ray method or the spatial partition binary tree method to distinguish the grid units inside the ore body model and the grid units outside the ore body model, and obtain an indication grid model.

5. The method for generating a local anisotropic search ellipsoid based on 3D-SOBEL according to claim 4, wherein Obtaining the boundary information of the ore body grid model includes: Use the 3D-SOBEL algorithm to calculate the ore body grid model to obtain a direction template, and perform a convolution calculation on the direction template and the indication grid model to obtain the boundary information of the ore body grid model.

6. The method for generating a local anisotropic search ellipsoid based on 3D-SOBEL according to claim 5, wherein The method for the convolution calculation is: Gx(x,y,z) = [f(x + 1,y - 1,z + 1)+2f(x + 1,y,z + 1)+f(x + 1,y + 1,z + 1)] - [f(x - 1,y - 1,z + 1)+2f(x - 1,y,z + 1)+f(x - 1,y + 1,z + 1)]+... Gy(x,y,z) = [f(x - 1,y - 1,z + 1)+2f(x,y - 1,z + 1)+f(x + 1,y - 1,z + 1)] - [f(x - 1,y + 1,z + 1)+2f(x,y + 1,z + 1)+f(x + 1,y + 1,z + 1)]+... Gz(x,y,z) = [f(x - 1,y - 1,z + 1)+2f(x - 1,y,z + 1)+f(x + 1,y + 1,z + 1)] - [f(x - 1,y - 1,z - 1)+2f(x - 1,y,z - 1)+f(x - 1,y + 1,z - 1)]+... Where f(x,y,z) is the indication value of the cuboid unit, and (x,y,z) is the center point coordinate.

Images