Thin vein ore body refined three-dimensional modeling method

Abstract

The invention relates to the technical field of three-dimensional geologic modeling, in particular to a fine three-dimensional modeling method for a thin-vein ore body. According to the method, an ore body contour line sequence and an ore body outer boundary line are generated through vectorization of geological maps such as a given exploration line profile map, line segment nodes of the ore body outer boundary line are encrypted through natural cubic spline function interpolation, and then corresponding biharmonic spline functions are fitted respectively; and adaptive dynamic interpolation is carried out on scanning lines in the sequence based on typical features of the thin-vein ore body. The curved surface meeting the specific smoothness condition is generated, it is ensured that the morphological change of the model is more reasonable, meanwhile, the interpenetration phenomenon is avoided, detailed features of areas such as faults and boundaries are effectively recognized, the accuracy of the model and the authenticity of geological expression are improved, and the method is suitable for popularization and application. Therefore, a more reliable thin-vein ore body refined three-dimensional modeling method is provided for geological exploration and resource evaluation.

Description

Technical Field

[0001] The present invention relates to the technical field of three-dimensional geological modeling, and particularly to a refined three-dimensional modeling method for thin vein-shaped ore bodies. Background Art

[0002] The three-dimensional geological model of an ore body is a closed surface formed by a top surface and a bottom surface in three-dimensional space. For thin vein-shaped ore bodies, the modeling is mainly based on the closed ore body contour lines vectorially drawn by geological workers and the outer boundary lines of the ore body. These closed contour lines are split into two open contour lines along the endpoints and are respectively defined as the upper boundary line and the lower boundary line of the ore body. The upper and lower boundary lines are combined with the outer boundary line to form the data of the top surface and the bottom surface of the ore body.

[0003] Traditional modeling techniques first increase the data density of the top surface and the bottom surface through methods such as discrete smooth interpolation (DSI) under the constraint of the outer boundary line of the ore body, and then use Delaunay triangulation to generate the top surface and the bottom surface of the model. However, for thin vein-shaped ore bodies, the interpolation and densification process may locally cause the interpenetration of the data of the top surface and the bottom surface, resulting in the intersection of the top surface and the bottom surface of the model and unable to accurately reflect the three-dimensional shape of the thin vein-shaped ore body. Therefore, traditional modeling techniques have limitations in constructing the three-dimensional model of thin vein-shaped ore bodies.

[0004] The three-dimensional modeling method based on ore body increment simulation first uses traditional modeling techniques to establish the bottom surface of the model. On this basis, by calculating the elevation difference between the lower boundary nodes and the corresponding upper boundary nodes of the ore body, an elevation increment function is fitted. The elevation increment of the bottom surface of the model is determined by using this function, and the bottom surface is correspondingly raised to form a new top surface. This method effectively avoids the intersection problem of the top surface and the bottom surface in the process of modeling thin vein-shaped ore bodies. However, this method assumes that the lower boundary and the upper boundary data are strictly vertically corresponding, while in fact, the ore body contour line nodes often distribute along an inclined direction and are not completely perpendicular to the Z-axis. This inclined distribution may cause errors in the calculation of the elevation difference, affecting the accuracy of the model, thus limiting the universality of this method.

[0005] Delaunay triangulation is an algorithm for three-dimensional surface visualization, which constructs a triangular mesh by connecting the given node data. The goal of this algorithm is to generate as regular triangles as possible because regular triangles contribute to generating a smoother three-dimensional model surface.

[0006] However, in actual modeling, the obtained ore body contour line node data is usually arranged directionally in space. When existing modeling methods encrypt data, they often do not fully consider the importance of the direction consistency between the new data and the original data. When the direction difference between the contour line node and its adjacent encrypted node is large, the shape difference between the generated triangular surface and the regular triangle will also increase, resulting in a decline in the quality of the triangular mesh in this area.

[0007] Therefore, it is necessary to provide a refined three-dimensional modeling method for thin vein-shaped ore bodies that can ensure the geometric accuracy of the model, improve the geological expression ability and visual quality of the model, and thus provide a more reliable method for geological exploration and resource assessment. Summary of the Invention

[0008] To overcome the problems of the prior art, the present invention proposes a refined three-dimensional modeling method for thin vein-shaped ore bodies that can ensure the geometric accuracy of the model, improve the geological expression ability and visual quality of the model, and thus provide a more reliable method for geological exploration and resource assessment.

[0009] A refined three-dimensional modeling method for thin vein-shaped ore bodies includes the following steps: Step 1: Based on the given exploration information, vectorize to generate a sequence of ore body contour lines and the outer boundary line of the ore body, and perform node encryption to generate the top interface data and bottom interface data of the model; Step 2: Based on the top interface data and bottom interface data of the model, respectively fit the corresponding biharmonic spline functions; Step 3: Project the top interface data and bottom interface data of the model onto the horizontal plane respectively, then generate a sequence of boundary-limited scan lines by the scan line method under the range constraint of the outer boundary line of the ore body, and finally perform adaptive dynamic interpolation on the scan lines in the sequence to generate two-dimensional directionally interpolated and encrypted nodes within the range of the top and bottom interfaces; Step 4: Determine the connection relationship between the nodes in the top interface and bottom interface data of the model; Step 5: Merge the top interface data and bottom interface data of the model, and splice the corresponding triangular mesh node connection matrix to form a seamless three-dimensional model of the ore body.

[0010] Further, Step 1 further includes: splitting the sequence of ore body contour lines into a sequence of upper boundary lines of the ore body and a sequence of lower boundary lines of the ore body, and sequentially interpolating and encrypting the line segment nodes of the sequence of upper boundary lines of the ore body, the sequence of lower boundary lines of the ore body, and the outer boundary line of the ore body through natural cubic spline functions.

[0011] Further, Step 1 further includes: combining the sequence of upper boundary lines of the ore body after node encryption with the outer boundary line of the ore body to form the top interface data of the model, and combining the sequence of lower boundary lines of the ore body after node encryption with the outer boundary line of the ore body to form the bottom interface data of the model.

[0012] Further, Step 2 further includes: Based on the top interface data and bottom interface data of the model, respectively call the surface fitting function Calculate the corresponding biharmonic spline functions and reconstruct the three-dimensional regular grid surface.

[0013] Further, step three further includes: the adaptive dynamic interpolation determines the encryption interpolation density of the scanning line based on the physical property parameters of the ore body and the data quality information.

[0014] Further, calculate the data confidence based on the multiple linear regression model; The multiple linear regression model is expressed as follows: Among them, 、 、 、 are regression coefficients, is the sampling accuracy, is the representativeness degree of the sampling data, is the spatial coverage, is the sampling repeatability error, which is divided into N regions according to the distance from the outer boundary line of the ore body, and the regression coefficients are determined for each region.

[0015] Further, the encryption interpolation density of the scanning line is calculated based on the physical property parameters of the ore body and the data confidence by parallel depth random forest, where in the multi-granularity scanning stage, the reduced feature set is filled and random sampling is performed on the feature subsequences after sliding window scanning.

[0016] Further, step three also includes that for the edge encryption nodes, that is, the head and tail nodes of the boundary-limiting scanning line, the elevation value is obtained by the method of assigning values according to the length ratio of the horizontal projection plane major axis of its connection with the segmentation line. The elevation value of the edge encryption node The calculation formula is as follows: 。

[0017] Further, step four also includes assigning the elevation values of the data of the top interface and the bottom interface of the model to 0, and using the two-dimensional boundary-limiting Delaunay triangulation function to generate the corresponding triangular connection matrices respectively; according to the data of the top interface and the bottom interface of the model and the triangular connection matrices, the top interface and the bottom interface of the model are visually displayed through triangular meshes.

[0018] Further, step five also includes: Directly splice the data sets and the triangular connection matrices of the top interface and the bottom interface of the model respectively to generate a combined data set and a combined triangular connection matrix, and then generate a fine three-dimensional model of a thin vein-shaped ore body with seamless splicing of the top and bottom interfaces of the model through triangular mesh visualization.

[0019] The present invention has the following advantages: 1. By introducing the biharmonic spline function, the present invention can generate a surface that meets specific smoothness conditions, which helps to ensure that the morphological changes of the model are more reasonable, avoid interpenetration phenomena, and improve the accuracy of the model and the authenticity of geological expression.

[0020] 2. The present invention uses a scan line algorithm to propose a method for generating an adaptive dynamic interpolation data set along a direction consistent with given ore body data. This method performs dynamic interpolation based on the typical characteristics of thin vein-shaped ore bodies. On the one hand, it can improve the quality of the triangulation network of the local model, making the surface of the three-dimensional ore body model smoother, enhancing the visual effect and analysis accuracy of the model. On the other hand, it can effectively identify the detailed characteristics of areas such as faults and boundaries. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 It is a schematic diagram of the directional interpolation process of the scan line method of the present invention.

[0022] Figure 2 It is a schematic diagram of the unencrypted ore body contour line and the outer boundary line of the ore body of the present invention.

[0023] Figure 3 It is a schematic diagram of the encrypted ore body contour line and the outer boundary line of the ore body of the present invention.

[0024] Figure 4 It is an X-Y horizontal projection diagram of the biharmonic spline surface corresponding to the top and bottom boundaries of the model of the present invention.

[0025] Figure 5 It is a schematic diagram of the relationship between the scan line direction and the scan progress direction of the present invention.

[0026] Figure 6 It is a schematic diagram of the relationship between the in-situ scan line and the interpolated scan line of the present invention.

[0027] Figure 7 It is a schematic diagram of the directional encryption of the boundary-defined scan line sequence of the present invention.

[0028] Figure 8 It is a schematic diagram of the three-dimensional directional node encryption of the top and bottom boundaries of the model of the present invention.

[0029] Figure 9 It is a schematic diagram of the three-dimensional visualization effect of the bottom boundary of the model of the present invention.

[0030] Figure 10 It is a schematic diagram of the effect of the fine three-dimensional model of the thin vein-shaped ore body of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0031] The following further clarifies the present invention in conjunction with specific embodiments. These embodiments are only used to illustrate the present invention and not to limit the scope of the present invention.

[0032] A method for fine three-dimensional modeling of a thin vein-shaped ore body includes: Step 1: First, based on given geological maps such as exploration line profile maps, manually vectorize them to generate a sequence of ore body contour lines and the outer boundary line of the ore body. Secondly, split the sequence of ore body contour lines into a sequence of upper boundary lines of the ore body and a sequence of lower boundary lines of the ore body.

[0033] On this basis, the sequence of upper boundary lines of the ore body, the sequence of lower boundary lines of the ore body, and the outer boundary line of the ore body are successively interpolated and encrypted for their line segment nodes through natural cubic spline functions. Among them, the sequence of upper boundary lines of the ore body after node encryption and the outer boundary line of the ore body are combined into the model top interface data; the sequence of lower boundary lines of the ore body after node encryption and the outer boundary line of the ore body are combined into the model bottom interface data.

[0034] Step 2: Based on the model top interface data and the model bottom interface data, fit the corresponding biharmonic spline functions respectively.

[0035] Step 3: Project the top and bottom interface data onto the horizontal plane respectively, then generate a sequence of boundary-limited scan lines by the scan line method under the range constraint of the outer boundary line of the ore body, and finally perform adaptive dynamic interpolation on the scan lines within the sequence, so as to generate two-dimensional directionally interpolated and encrypted nodes within the range of the top and bottom interfaces.

[0036] As Figure 1 shown, for the above two-dimensional directionally interpolated and encrypted nodes, use the biharmonic spline function of the model top interface and the biharmonic spline function of the model bottom interface fitted in Step 2 to assign elevation values to them respectively, and generate three-dimensional directionally encrypted point sets for the model top interface and the bottom interface.

[0037] Step 4: First, clean the model top interface and bottom interface data, delete redundant node data of the outer boundary line of the ore body, and regenerate the outer boundary line of the ore body using the optimized data.

[0038] Secondly, assign the elevation values of the model top interface data and the model bottom interface data to 0, project them onto the two-dimensional X-Y plane, and then use the Delaunay triangulation criterion under boundary constraints with the outer boundary line of the ore body as the boundary limit condition to generate the corresponding triangular mesh node connection matrices respectively, and determine the connection relationships between the nodes in the model top interface and bottom interface data.

[0039] Therefore, according to the triangular mesh node connection matrices corresponding to the model top interface and bottom interface data respectively, the model top interface and the model bottom interface of the ore body can be visually displayed using the triangular mesh.

[0040] Step 5: Since the top interface of the ore body and the bottom interface of the ore body use the new outer boundary line of the ore body as the common boundary, merge the model top interface and bottom interface data, and splice the corresponding triangular mesh node connection matrices, then a seamless three-dimensional model of the ore body can be formed. This model can more accurately express the morphological changes of the ore body and has a high-quality surface.

[0041] Detailed process of Step 1: A1. Import geological maps such as exploration line profile maps into modeling software such as Micromine and perform coordinate correction, and then generate a vectorized ore body contour line sequence by manually drawing the ore body contour line on the profile.

[0042] Based on the relative orientation of the ore body contour line nodes and combined with the experience of geological personnel, infer the positions of the outer boundary points of the ore body strike, and then sequentially connect the end points on both sides of the ore body contour line and the strike boundary points to generate a vectorized outer boundary line of the ore body.

[0043] A2. For the ore body contour line sequence, divide it into an upper boundary line sequence and a lower boundary line sequence of the ore body by splitting along the end points on both sides of each internal contour line.

[0044] A3. Import the line segment node coordinate data of the upper boundary line sequence, lower boundary line sequence, and outer boundary line of the ore body into MatLab.

[0045] For each upper boundary line, lower boundary line, and outer boundary line of the ore body, fit a natural cubic spline function using their line segment node coordinates , thereby generating the corresponding spline curve , and then according to the curve drawing function obtain the control points of the corresponding spline curve as encrypted nodes. As encrypted nodes.

[0046] The encrypted upper boundary line sequence and outer boundary line of the ore body form the top interface data of the model; the encrypted lower boundary line sequence and outer boundary line of the ore body form the bottom interface data of the model.

[0047] As Figure 2 and Figure 3 shown, it can be seen the comparison of the effects before and after encryption of the upper boundary line sequence, lower boundary line sequence, and outer boundary line of the ore body. The encrypted ore body contour line and outer boundary line of the ore body are smoother, reducing geometric errors and visual defects.

[0048] Detailed process of Step 2: A1. Based on the top interface data and bottom interface data of the model, respectively call the surface fitting function built in MATLAB to calculate the corresponding biharmonic spline function.

[0049] A2. Reconstruct a three-dimensional regular grid surface based on the biharmonic spline functions corresponding to the top and bottom boundaries, and project it onto the horizontal plane.

[0050] From Figure 4It can be seen that the top and bottom interfaces share the outer boundary line of the ore body as their common boundary. The boundary between the two is distinct, and there is no interpenetration phenomenon inside. A surface that meets specific smoothness conditions can be generated, which helps to ensure that the morphological changes of the model are more reasonable, avoid interpenetration phenomena, and improve the accuracy of the model and the authenticity of geological expression.

[0051] The detailed process of Step 3 is as follows: A1. Project the data of the top interface of the model onto the horizontal plane. The horizontal projection line segment sequence of the upper boundary line of the ore body is a set of parallel line segments. Take the line segment direction vector as the scanning line direction vector v. Given that the projection result of the upper boundary line sequence of the ore body is a series of parallel line segments, the scanning line direction vector is calculated by the following formula v : In the formula, The horizontal coordinate of the first node of the upper boundary line of the ore body; The coordinates of the last node of the upper boundary line of the ore body.

[0052] A2. Let the scanning line sequence travel in a direction perpendicular to the scanning line, that is, the scanning line travel direction vector Then it is calculated by the following formula :

[0053] In the above formula, define , that is, the scanning line sequence travels along the positive direction of the axis, and is obtained, so as to obtain the initial value of the scanning line travel direction vector . As Figure 5 shown, the relationship between the scanning line direction vector and the scanning line travel direction vector can be seen.

[0054] A3. Based on the horizontal projection line segment sequence of the upper boundary line of the ore body, obtain the boundary-limiting scanning line sequence according to the following steps: ① Extend the head and tail nodes of each line segment in the horizontal projection line segment sequence of the upper boundary line of the ore body along the scanning line direction, and take the connection line of the extended head and tail nodes of each line segment to form the in-situ scanning line sequence. Among them, the calculation formula for the extended coordinates of the head and tail nodes of the horizontal projection line segment of the upper boundary line of the ore body is as follows: In the formula, ② Assume that the head and tail boundary point coordinates of any in-situ scanning line are respectively , and the head node coordinates of the adjacent in-situ scanning line on the side of the scanning line travel direction are , then the distance between the two is calculated by the following formula . In the formula, Determinant The absolute value of. ③ Given the distance between adjacent in-situ scan lines , set the scanning line travel spacing as , and calculate the number of scanning lines generated by the current in-situ scan line along the scanning travel direction according to the following formula . Then, according to the number of scanning lines calculated by the above formula , recalculate the actual scanning line travel spacing : Starting from the current in-situ scan line, along the scanning line travel direction, at a spacing of generate directed scanning lines. The coordinates of the head and tail nodes of the scanning lines can be calculated by the following formula: In the formula, The tail node coordinates of the i-th scanning line; The tail node coordinates of the i-th scanning line.

[0055] Repeat this step for each pair of adjacent in-situ scan lines to generate a scanning line sequence. As Figure 6 shown, the relationship between the in-situ scan line and the interpolated scan line sequence is rotated counterclockwise by 20° relative to Figure 4 .

[0056] ④ Sort the node coordinates of the projection line segments of the outer boundary line of the ore body separately by the X-axis and Y-axis to obtain the maximum and minimum values of the coordinates and the coordinates. Secondly, split the projection line segments of the outer boundary line of the ore body into pairs of two nodes each to generate a sequence of segmented outer boundary lines.

[0057] For each scanning line in the scanning line sequence, by traversing and searching the sequence of segmented outer boundary lines, calculate the cross product of the connection vector between the head and tail nodes of each segmented outer boundary line and the connection vector between its head and tail nodes. The cross product calculation formula is as follows: The cross product calculation formula is as follows: In the formula, The head and tail nodes of the segmented boundary line; The head and tail nodes of the scanning line.

[0058] On this basis, follow the following principle to obtain the boundary-limiting scanning line: When the cross product of the directed scanning line sequence and two or more segmented outer boundary lines is not 0, then there are two or more intersection points between the directed scanning line and the outer boundary line of the ore body. If the midpoint of the intersection point connection satisfies and at the same time, then the connection line between the two intersection points is the boundary-limiting scanning line. The formula for calculating the intersection point coordinates is as follows: Based on the above principle, according to the cross product results of the connection vectors between the head and tail nodes of each scanning line and the segmented outer boundary lines, obtain the sequence of boundary-limiting scanning lines.

[0059] A4. As Figure 7 shown, based on the boundary-defined scan line sequence, segment nodes of each scan line are encrypted to generate an oriented encrypted point set on the projection interface. For edge encrypted nodes, i.e., the start and end nodes of the boundary-defined scan line, elevation values are obtained by the method of assigning values according to the length ratio of the segment boundary line along the major axis of the horizontal projection plane, so as to ensure the consistency of the top and bottom interface boundaries of the model. The elevation values of the edge encrypted nodes are calculated as follows: Then, elevation values are assigned to the internal oriented encrypted point set by using the top interface biharmonic spline function and the bottom interface biharmonic spline function respectively, and a three-dimensional oriented encrypted point set of the model top interface and the bottom interface as Figure 8 shown can be obtained.

[0060] Optionally, the present invention dynamically determines the scan line encryption interpolation density based on key ore body data to realize dynamic interpolation of geological features adapting to themselves. Since the typical characteristics of thin vein-shaped ore bodies are small thickness, unstable extension, and large grade fluctuation, in the existing fine three-dimensional modeling methods for thin vein-shaped ore bodies, conventional interpolation methods usually adopt the way of uniform interpolation, which cannot effectively identify the detailed features of regions such as faults and boundaries, and cannot adjust the modeling parameters according to the reliability of sampling. Based on the problems existing in the above-mentioned prior art, the present invention dynamically determines the scan line encryption interpolation density based on key ore body data, which specifically includes the following steps: a. Obtain the physical property parameters of the ore body and data quality information; The physical property parameters of the ore body include the ore grade distribution gradient, the distance from the outer boundary line of the ore body, the ore body thickness change rate, and the structural curvature; Among them, the ore grade distribution gradient refers to the change intensity of the content of useful elements in the ore in space, and its core feature is the significant mutation of the grade within a unit distance, which is mathematically manifested as the region of the maximum gradient value of the grade function; The data quality information includes sampling accuracy, such as drilling density, and the representativeness of sampling data, such as core recovery rate, spatial coverage, and sampling repeatability error.

[0061] b. Determine the data confidence level according to the data quality information.

[0062] Optionally, calculate the data confidence level based on a multiple linear regression model; The multiple linear regression model is expressed as follows: Among them, , , , are regression coefficients, is the sampling accuracy, is the representativeness of sampling data, is the spatial coverage is the sampling repeatability error. To achieve refined modeling of the thin-vein ore body, according to the characteristics of the thin-vein ore body, it is divided into N regions according to the distance from the outer boundary line of the ore body, and the regression coefficients are determined for each region.

[0063] Optionally, the regression coefficients are estimated based on the least squares method.

[0064] c. Calculate the encrypted interpolation density of the scan line based on the physical property parameters of the ore body and the data confidence level using parallel deep random forest.

[0065] Includes feature dimensionality reduction, multi-granularity scanning, cascade forest construction, and load balancing stages; Optionally, to ensure the smoothness and balance of the data, the multi-granularity scanning stage of the present invention fills the reduced feature set and randomly samples the feature subsequences after sliding window scanning.

[0066] Compared with the traditional deep neural network structure, parallel deep random forest has the following advantages: In terms of algorithm performance, it is higher than the traditional deep neural network. It has easy trainability, significantly fewer parameters required than the deep neural network, the results of the algorithm are less sensitive to the parameter settings, the hyperparameter settings have extremely high robustness, and it does not require training with a large amount of data. Even on small data samples, it can have good performance and relatively accurate experimental results; the algorithm efficiency is relatively high and has scalability. It is suitable for parallel deployment, can achieve efficient parallel training, and does not require large-scale manual parameter setting and tuning like the prediction method of the neural network.

[0067] The detailed process of step four: A1. For the data of the top and bottom interfaces of the model after directional encryption, delete the redundant line segment nodes in the outer boundary line of the ore body, that is, the isolated nodes without adjacent nodes along the scan line direction.

[0068] A2. Reconnect the line segment nodes of the outer boundary line of the ore body and the edge encryption nodes in sequence within the data of the top interface of the model to obtain a new outer boundary line of the ore body.

[0069] A3. Assign the elevation values of the data of the top and bottom interfaces of the model to 0 again, and then use the two-dimensional boundary-defined Delaunay triangulation function in MatLab to generate the corresponding triangular connection matrices respectively. As Figure 9 shown, and then according to the data of the top and bottom interfaces of the model and the triangular connection matrices, directly visualize and display the top interface and the bottom interface of the model through triangular meshes.

[0070] The detailed process of step five: AsFigure 10 As shown, the datasets of the top and bottom interfaces of the model and the triangular connection matrix are directly spliced respectively to generate a combined dataset and a combined triangular connection matrix. Then, through triangular mesh visualization, a fine three-dimensional model of the thin-vein-shaped ore body with seamless splicing of the top and bottom interfaces of the model can be generated. The present invention can capture and accurately express the morphological changes of the ore body. Whether it is a complex geometric shape or a subtle geological feature, it can be effectively presented. Moreover, the application is not limited to a specific type of ore body. Whether it is a thick and large ore body or a thin-vein-shaped ore body, an effective three-dimensional modeling can be obtained through the method of the present invention, showing its wide applicability.

[0071] The above embodiments have described in detail the preferred embodiments of the present invention. However, the present invention is not limited thereto. Within the scope of the technical concept of the present invention, various simple modifications can be made to the technical solutions of the present invention, including any other suitable combination of each technical feature. These simple modifications and combinations should also be regarded as the content disclosed by the present invention and fall within the protection scope of the present invention.

Claims

1. A refined three-dimensional modeling method for thin vein-shaped ore bodies, characterized in that, It includes the following steps: Step 1: Based on the given exploration information, vectorize to generate a sequence of ore body contour lines and the outer boundary line of the ore body, and perform node encryption to generate the data of the top interface and the bottom interface of the model; Step 2: Based on the data of the top interface and the bottom interface of the model, fit the corresponding biharmonic spline functions respectively; Step 3: Project the data of the top interface and the bottom interface of the model onto the horizontal plane respectively, then generate a sequence of boundary-limiting scan lines by the scan line method under the range constraint of the outer boundary line of the ore body, and finally perform adaptive dynamic interpolation on the scan lines in the sequence to generate two-dimensional directional interpolation encrypted nodes within the range of the top and bottom interfaces; Step 4: Determine the connection relationship between the nodes in the data of the top interface and the bottom interface of the model; Step 5: Merge the data of the top interface and the bottom interface of the model, and splice the corresponding triangular mesh node connection matrix to form a seamless three-dimensional model of the ore body.

2. The refined three-dimensional modeling method for a thin vein-shaped ore body according to claim 1, wherein The said Step 1 further includes: splitting the sequence of ore body contour lines into a sequence of upper boundary lines of the ore body and a sequence of lower boundary lines of the ore body, and interpolating and encrypting the line segment nodes of the sequence of upper boundary lines of the ore body, the sequence of lower boundary lines of the ore body, and the outer boundary line of the ore body in turn through natural cubic spline functions.

3. A refined three-dimensional modeling method for thin vein-shaped ore bodies according to claim 2, characterized in that The said Step 1 further includes: combining the sequence of upper boundary lines of the ore body after node encryption and the outer boundary line of the ore body to form the data of the top interface of the model, and combining the sequence of lower boundary lines of the ore body after node encryption and the outer boundary line of the ore body to form the data of the bottom interface of the model.

4. A refined three-dimensional modeling method for thin vein-shaped ore bodies according to claim 1, characterized in that The said Step 2 further includes: Based on the model top interface data and the model bottom interface data, respectively call the surface fitting function Calculate the corresponding biharmonic spline function and reconstruct the three-dimensional regular grid surface.

5. A refined three-dimensional modeling method for thin vein-shaped ore bodies according to claim 1, characterized in that, The said Step 3 further includes: the adaptive dynamic interpolation determines the scan line encryption interpolation density based on the physical property parameters of the ore body and the data quality information.

6. The refined three-dimensional modeling method for a thin vein-shaped ore body according to claim 5, wherein, Calculate the data confidence based on the multiple linear regression model; The multiple linear regression model is expressed as follows: where , , , are regression coefficients, is the sampling accuracy, is the representativeness degree of the sampling data, is the spatial coverage, is the sampling repeatability error, which is divided into N regions according to the distance from the outer boundary line of the ore body, and the regression coefficients are determined for each region.

7. A refined three-dimensional modeling method for thin vein-shaped ore bodies according to claim 5, characterized in that The scan line encryption interpolation density is calculated based on the physical property parameters of the ore body and the data confidence by parallel deep random forest, where in the multi-granularity scanning stage, the dimensionality-reduced feature set is filled and random sampling is performed on the feature subsequences after sliding window scanning.

8. A refined three-dimensional modeling method for thin vein-shaped ore bodies according to claim 7, characterized in that Step 3 further includes, for the edge encryption nodes, that is, the head and tail nodes of the boundary-defining scanning line, obtaining the elevation value by the method of assigning values according to the length ratio of the horizontal projection plane major axis between it and the segmentation boundary line, and the elevation value of the edge encryption node The calculation formula is as follows: In the formula, (x1, y1) is the horizontal coordinate of the head node of the upper boundary line of the ore body; (x2, y2) is the coordinate of the tail node of the upper boundary line of the ore body, and x mid is the abscissa of the midpoint of the connection line of the intersection points of the directional scanning line and the outer boundary line of the ore body.

9. A refined three-dimensional modeling method for thin vein-shaped ore bodies according to claim 1, characterized in that Step 4 further includes assigning an elevation value of 0 to the elevation values of the top and bottom interface data of the model, and using a two-dimensional boundary-defined Delaunay triangulation function to generate corresponding triangular connection matrices respectively; according to the top and bottom interface data of the model and the triangular connection matrices, the top and bottom interfaces of the model are visually displayed through triangular meshes.

10. A refined three-dimensional modeling method for thin vein-shaped ore bodies according to claim 1, characterized in that The said Step 5 further includes: Directly splice the data sets of the top interface and the bottom interface of the model and the triangular connection matrix respectively to generate a merged data set and a merged triangular connection matrix, and then generate a fine three-dimensional model of the thin vein-shaped ore body with seamless splicing of the top and bottom interfaces of the model through triangular mesh visualization.

Images