Multi-ore body contour line modeling method, electronic device and storage medium

Abstract

The application discloses a multi-ore body contour line modeling method, an electronic device and a storage medium. The method comprises the following steps: determining a set of plane segmentation lines of each profile in adjacent profiles based on preset threshold parameters and a plurality of ore body contour lines corresponding to each profile in the adjacent profiles; constructing an internal space segmentation domain between narrow plane segmentation lines of adjacent two profiles and an external space segmentation domain between overlapping plane segmentation lines or gap plane segmentation lines of the adjacent two profiles based on an optimal matching pair set; extracting a shortest connection path between a start anchor point and an end anchor point of a section based on a dynamic programming algorithm with a forbidden point set, the shortest connection path not intersecting with the internal space segmentation domain and the external space segmentation domain; generating a connection triangular net of each ore body based on each shortest connection path; and constructing a multi-ore body three-dimensional wireframe model based on each connection triangular net. In this way, the multi-ore body contact position is completely fitted, and the problems of distortion and self-intersection in the modeling of a narrow ore body are solved.

Description

Technical Field

[0001] This application relates to the field of 3D modeling, and in particular to a method for modeling the contours of multiple ore bodies, electronic devices, and storage media. Background Technology

[0002] In mineral exploration and resource assessment, it is often necessary to construct a three-dimensional wireframe model of the ore body based on the ore body outlines interpreted from adjacent exploration line profiles. This model is used for reserve estimation, mining design, and visualization analysis. When multiple ore bodies exist on a profile, their topological relationships in three-dimensional space are complex. They may be in contact with each other, partially coplanar, or, although not in direct contact, extremely close in spatial distance.

[0003] Currently, when dealing with the connection of multiple ore body contours between adjacent profiles, the common strategy is to independently connect the contour lines of each ore body using triangular meshes. That is, for each ore body, its corresponding contour on the adjacent profile is selected separately, and an independent surface triangular mesh of the ore body is generated using algorithms such as the shortest diagonal method, the maximum volume method, or the dynamic programming method. However, the above independent modeling strategy has the following technical problems in practical engineering applications: (1) The models do not fit at overlapping and coplanar locations. When two ore bodies are in contact in space and share a part of the boundary (i.e., coplanar), since each ore body independently constructs a triangular mesh, it is difficult to maintain a strict consistency in the triangular mesh connection relationship, vertex distribution, and normal direction in the shared boundary area. In the final generated three-dimensional model, gaps or mutual penetrations often appear between the triangular meshes at the contact position, which cannot truly reflect the geological characteristics of the ore body contact surface and seriously affects the accuracy of subsequent volume calculations. (2) Model interference at adjacent gaps. When two ore bodies are not in contact in space but are very close to each other (such as in narrow veins or branching composite parts), independent modeling lacks a spatial constraint mechanism between them. When the triangular meshes on the surface of each ore body approach their respective contour boundaries, they are prone to expand outward or deviate, thus encroaching on the spatial area occupied by the adjacent ore body. This non-physical model interference not only violates geological common sense, but also causes model-based spatial analysis (such as distance detection and grade interpolation) to produce incorrect results. (3) Connection distortion and self-intersection of narrow ore bodies. For ore bodies with narrow shapes such as veins and lenses, their contours often have subtle changes in shape and orientation on adjacent profiles. When traditional independent connection algorithms deal with such narrow contours, due to the lack of forced guidance constraints on the connection direction, the generated triangular meshes are very prone to distortion and deformation in narrow areas, and even the phenomenon of triangular facets folding and intersecting (i.e., self-intersection) occurs. Especially when two adjacent ore bodies are both narrow and close in space, the risk of intersection caused by independent connection will increase sharply, leading to complete model failure.

[0004] In summary, existing methods cannot uniformly identify and constrain the overlapping coplanar boundaries between ore bodies, adjacent gap regions, and the elongated structure of the ore body itself before or during modeling. This results in a lack of rigid constraints in the subsequent triangulation generation process, making it difficult to guarantee the fit of the multi-ore body 3D model at the contact interface, the mutual exclusion of spatial distribution, and the smoothness of the elongated parts. Summary of the Invention

[0005] In view of this, embodiments of this application provide a multi-orebody contour modeling method, electronic device, and storage medium, aiming to provide a multi-orebody contour modeling scheme that automatically identifies overlaps, gaps, and elongated structures between orebody contours and constructs spatial segmentation domains as constraints, enabling the collaborative generation of surface models of each orebody under these constraints. This solves the technical problems in the prior art caused by the lack of spatial constraint mechanisms between orebody 3D models, such as non-fitting at contact interfaces, mutual interference between adjacent spaces, and twisted self-crossing in elongated parts.

[0006] The technical solution of this application embodiment is implemented as follows:

[0007] In a first aspect, embodiments of this application provide a method for modeling the contour lines of multiple ore bodies, the method comprising:

[0008] Obtain multiple ore body contour lines and preset threshold parameters on adjacent cross sections, wherein the multiple ore body contour lines are a set of points arranged in counterclockwise order;

[0009] Based on the preset threshold parameters and the multiple ore body contour lines corresponding to each profile in the adjacent profiles, a set of planar dividing lines for each profile in the adjacent profiles is determined; wherein, the preset threshold parameters include: overlap determination tolerance, elongated shape factor threshold, elongated width threshold, and gap identification threshold; the set of planar dividing lines includes: overlapping planar dividing lines, elongated planar dividing lines, and gap planar dividing lines; the set of planar dividing lines includes a first set of planar dividing lines for a first profile and a second set of planar dividing lines for a second profile adjacent to the first profile;

[0010] The first set of planar segmentation lines and the second set of planar segmentation lines are respectively used as the first planar segmentation sequence and the second planar segmentation sequence. A similarity matrix is ​​established between the first planar segmentation sequence and the second planar segmentation sequence. The maximum cumulative similarity function constructed by the similarity matrix is ​​solved based on dynamic programming, and the optimal matching pair set is generated based on the path backtracking corresponding to the maximum cumulative similarity. Based on the optimal matching pair set, an internal spatial segmentation domain is constructed between the narrow planar segmentation lines of two adjacent profiles and an external spatial segmentation domain is constructed between the overlapping planar segmentation lines or gap planar segmentation lines of two adjacent profiles.

[0011] The endpoints of each element in the optimal matching pair set are used as forced anchor points on the ore body contours of the first and second profiles. Based on the forced anchor points, each ore body contour is divided into multiple one-to-one corresponding segments. For each of the multiple segments, with the internal spatial segmentation domain and the external spatial segmentation domain as hard constraints, a dynamic programming algorithm with a prohibited point set is used to extract the shortest connection path between the starting anchor point and the ending anchor point of the segment that does not intersect with the internal spatial segmentation domain and the external spatial segmentation domain. Based on each of the shortest connection paths, a connection triangular network of each ore body is generated. Based on each of the connection triangular networks, a three-dimensional wireframe model of multiple ore bodies is constructed.

[0012] In some embodiments, determining the set of planar dividing lines for each profile in the adjacent profiles based on the preset threshold parameter and the multiple ore body contour lines corresponding to each profile in the adjacent profiles includes:

[0013] Based on the overlap determination tolerance, the overlapping plane dividing line between multiple ore body outlines in each of the adjacent profiles is determined;

[0014] Based on the elongated shape factor threshold and elongated width threshold, elongated planar dividing lines between multiple ore body outlines in each of the adjacent profiles are determined;

[0015] Based on the gap identification threshold, the gap plane dividing line between multiple ore body contours in each of the adjacent profiles is determined;

[0016] Based on the overlapping plane dividing line, the narrow plane dividing line, and the gap plane dividing line, the set of plane dividing lines is generated;

[0017] The step of determining the overlapping plane dividing line between multiple ore body contours in each of the adjacent profiles based on the overlap determination tolerance includes:

[0018] For each of the adjacent profiles, traverse all sub-line segment pairs of the multiple ore body contours within the profile;

[0019] Extract overlapping sub-segment pairs from all sub-segments that satisfy a first set condition and a second set condition; the first set condition is that the first perpendicular distance between the two vertices of the first sub-segment and the second sub-segment or its extension in the sub-segment pair is less than the overlap determination tolerance; the second set condition is that at least one vertex of the first sub-segment and the second sub-segment has a perpendicular distance less than the overlap determination tolerance; extract overlapping segments from the overlapping sub-segment pairs and merge consecutive overlapping segments; generate an overlapping plane segmentation line set for the profile based on each consecutive overlapping segment; generate the overlapping plane segmentation line set based on the overlapping plane segmentation line set for each profile.

[0020] The step of determining the plane dividing line of the gap between multiple ore body contours in each of the adjacent profiles based on the gap identification threshold includes:

[0021] For each of the adjacent profiles, traverse all boundary sub-segment pairs of multiple ore body contours within the profile, and extract the nearest neighbor segments between all boundary sub-segment pairs whose nearest point distance is less than the gap identification threshold.

[0022] Remove the areas where the adjacent line segments have been determined to overlap, and generate gap boundary segments;

[0023] Based on the narrow domain skeleton extraction algorithm, a triangular network is constructed for the gap boundary segment; based on the gap topological extension direction, the midpoints of all internal non-boundary edges in the triangular network are sequentially connected to generate the gap plane segmentation line.

[0024] In some embodiments, determining the elongated planar dividing line between multiple ore body contours in each of the adjacent profiles based on the elongated shape factor threshold and the elongated width threshold includes:

[0025] For each of the multiple orebody contours in each of the adjacent cross-sections, the overall shape factor of the orebody contour is obtained. If the overall shape factor is less than the elongated shape factor threshold, constrained triangulation is performed on the interior of the orebody contour to generate multiple constrained triangulation triangles. These multiple constrained triangulation triangles satisfy positional constraints, namely: among the multiple constrained triangulation triangles, only the triangulation triangles located at the two ends of the contour have two sides that fit the orebody contour, while all other internal triangles have only one side that fits the orebody contour. For each constrained triangulation triangle, the vertex opposite the edge that fits the orebody contour is used as the reference edge. The third perpendicular distance to the reference edge is determined. If the third perpendicular distance is less than the narrow width threshold, the triangle is determined to be a narrow triangle. All adjacent narrow triangles are connected to form a narrow domain of the ore body. Based on the topological relationship between the narrow domain of the ore body and the ore body outline, the type of the narrow domain of the ore body is determined. The types of the narrow domain include: fully enclosed narrow domain, semi-open narrow domain, and fully open narrow domain. The fully enclosed narrow domain contains two characteristic triangles, and the two sides of the two characteristic triangles fit the ore body outline. The semi-open narrow domain contains a unique characteristic triangle, and the two sides of the characteristic triangle fit the ore body outline. The fully open narrow domain does not contain a characteristic triangle whose two sides fit the ore body outline.

[0026] Based on the type of the narrow domain of the ore body, a target narrow domain is generated; a triangular mesh is constructed for the target narrow domain; the midpoints of all internal edges in the triangular mesh that do not fit with the outline of the ore body are extracted, and the midpoints are connected in an orderly manner according to the topological extension direction to construct the narrow plane dividing line of the outline of the ore body.

[0027] Based on the narrow planar segmentation lines of each of the ore body contour lines, the narrow planar segmentation lines of the profile are generated;

[0028] The elongated plane dividing lines are generated based on the elongated plane dividing lines of each of the aforementioned cross-sections.

[0029] In some embodiments, generating a target narrow region based on the type of the ore body narrow region includes:

[0030] If the type of the ore body narrow domain is a fully enclosed narrow domain or a semi-open narrow domain, then the ore body narrow domain is preprocessed to generate the target narrow domain;

[0031] If the type of the ore body elongated region is a fully open elongated region, then the fully open elongated region is the target elongated region;

[0032] Wherein, if the type of the narrow ore body domain is a fully enclosed narrow domain or a semi-open narrow domain, then the narrow ore body domain is preprocessed to generate the target narrow domain, including:

[0033] If the type of the ore body elongated domain is a fully enclosed elongated domain, then the common vertex of the two characteristic triangles with two sides that fit the outline of the ore body within the fully enclosed elongated domain is used as the cut-off point to decompose the closed elongated domain into two independent polylines that are not connected to each other; based on each of the independent polylines that are not connected to each other, the pre-processed elongated domain is generated.

[0034] If the type of the ore body elongated domain is a semi-open elongated domain, then the common vertex of the characteristic triangle whose only two sides in the semi-open elongated domain region conform to the topological features of the ore body outline is used as the cutoff point, and the semi-open elongated domain is decomposed into two independent polylines that are not connected to each other. Based on the two independent polylines that are not connected to each other, the pre-processed elongated domain is generated.

[0035] In some embodiments, the similarity matrix is:

[0036] ;

[0037] in, The i-th dividing line in the first planar segmentation sequence With the j-th dividing line in the second plane segmentation sequence Similarity between them For length similarity, For location proximity, The dot product of unit direction vectors. These are the weighting coefficients, This is the sequence of first plane dividing lines. This is the sequence of second-plane dividing lines;

[0038] The maximum cumulative similarity function is:

[0039] ;

[0040] in, From the starting position to the current position The maximum cumulative similarity value, The i-th dividing line in the first planar segmentation sequence With the j-th dividing line in the second plane segmentation sequence The similarity between them.

[0041] In some embodiments, the construction of an internal spatial segmentation domain between narrow planar dividing lines of adjacent two sections and an external spatial segmentation domain between overlapping planar dividing lines or gap planar dividing lines of adjacent two sections, based on the optimal matching pair set, includes:

[0042] If the ore bodies corresponding to the two adjacent profiles are both narrow ore bodies, then the matching pairs in the optimal matching pair set that correspond to the narrow plane dividing lines are taken as internal matching pairs. For each internal matching pair, the open segment wireframe algorithm is used to construct the two narrow plane dividing lines corresponding to the matching pair into a triangular mesh surface. The internal spatial segmentation domain is constructed based on each triangular mesh surface.

[0043] If there are overlapping plane dividing lines or gap plane dividing lines on both adjacent cross-sections, then the matching pairs in the optimal matching pair set corresponding to the overlapping plane dividing lines or gap plane dividing lines are taken as external matching pairs; for each external matching pair, the two dividing lines corresponding to the matching pair are connected to generate a triangular mesh surface; based on each triangular mesh surface, the external spatial segmentation domain is constructed.

[0044] In some embodiments, the extraction of the shortest connection path between the start and end anchor points of the segment that does not intersect with the internal and external spatial partitioning domains, based on the dynamic programming algorithm with a forbidden point set, includes:

[0045] For each segment, the starting anchor point and the ending anchor point of the segment are used as the endpoints that must be passed through, and the contour point sequence on the first profile and the contour point sequence on the second profile within the segment are used as the set of points to be matched.

[0046] A dynamic programming recursive function is constructed based on the point distance cost function and the forbidden point set. The point distance cost function is used to characterize the spatial distance between any contour point on the first profile and any contour point on the second profile. The forbidden point set is formed when a triangular facet generated by connecting any contour point on the first profile and any contour point on the second profile intersects with any of the spatial partitioning domains.

[0047] Based on the path function and the dynamic programming recursive function, the shortest connection path that does not intersect with the internal and external spatial partitioning domains is extracted between the starting and ending anchor points of the segment. The path function is a sequence of point pairs from the starting anchor point to the ending anchor point, satisfying monotonicity, continuity, and forcibly passing through the starting and ending anchor points. The dynamic programming recursive function is recursively calculated by comparing the cumulative costs of three paths to the current position from the diagonal, upward, and leftward directions, and includes a legal constraint: the legal constraint is that the newly generated triangular facets during the migration process do not intersect with any of the spatial partitioning domains, and the intermediate points are not within the forbidden point set Ω; if the migration in a certain direction does not satisfy the legal constraint, the cost of that direction is set to +∞.

[0048] In some embodiments, the path function is:

[0049] ;

[0050] in, Let be the path function, representing the optimal connection scheme from the starting anchor point to the ending anchor point. For the first on the path The nth point pair represents the nth point of the first profile. The first contour point and the second profile of the second section Connect the contour points The path point number. This represents the total number of path pairs;

[0051] The dynamic programming recursive function is:

[0052] ;

[0053] in, To start from the anchor point Go to current location The cumulative minimum cost value, For the cost of distance between points, From the starting point to The cumulative cost, From the starting point to The cumulative cost, From the starting point to The cumulative cost, , To obtain the minimum value among the three paths, a valid path is defined as one where the newly generated triangular facets during the migration process do not intersect with any spatial partitioning domain, and the intermediate points are not in the forbidden point set. Inside; otherwise, the value in that direction is set to... .

[0054] In a second aspect, embodiments of this application provide an electronic device, including: a processor and a memory for storing a computer program capable of running on the processor, wherein the processor, when running the computer program, performs the steps of the method described in the first aspect above.

[0055] Thirdly, embodiments of this application provide a computer storage medium storing a computer program, characterized in that, when the computer program is executed by a processor, it implements the steps of the method described in the first aspect above.

[0056] This application provides a method for modeling multiple ore body contours. The method includes: acquiring multiple ore body contours and preset threshold parameters on adjacent cross-sections, wherein the multiple ore body contours are a set of points arranged in a counter-clockwise order; determining a set of planar segmentation lines for each cross-section in adjacent cross-sections based on the preset threshold parameters and the multiple ore body contours corresponding to each cross-section in adjacent cross-sections; wherein the preset threshold parameters include: overlap judgment tolerance, elongated shape factor threshold, elongated width threshold, and gap identification threshold; the set of planar segmentation lines includes: overlapping planar segmentation lines, elongated planar segmentation lines, and gap planar segmentation lines; the set of planar segmentation lines includes a first set of planar segmentation lines for a first cross-section and a second set of planar segmentation lines for a second cross-section adjacent to the first cross-section; establishing a similarity matrix between the first and second planar segmentation sequences by using the first and second set of planar segmentation lines as a first planar segmentation sequence and a second planar segmentation sequence, respectively; and based on dynamic rules... The algorithm solves for the maximum cumulative similarity function constructed from the similarity matrix and backtracks to generate the optimal matching pair set based on the path corresponding to the maximum cumulative similarity. Based on the optimal matching pair set, it constructs an internal spatial segmentation domain between the narrow plane segmentation lines of two adjacent profiles and an external spatial segmentation domain between the overlapping plane segmentation lines or gap plane segmentation lines of two adjacent profiles. The endpoints of each element in the optimal matching pair set are used as forced anchor points on the ore body contours of the first and second profiles. Based on the forced anchor points, each ore body contour is divided into multiple one-to-one corresponding segments. For each segment, with the internal and external spatial segmentation domains as hard constraints, a dynamic programming algorithm with a forbidden point set is used to extract the shortest connection path between the starting and ending anchor points of the segment that does not intersect with the internal and external spatial segmentation domains. Based on each shortest connection path, a connection triangular network of each ore body is generated. Based on each connection triangular network, a three-dimensional wireframe model of multiple ore bodies is constructed.

[0057] Thus, given the existence of multiple ore body contours on adjacent exploration profiles, this application provides a collaborative modeling method for multiple ore body contours, which can automatically identify overlaps, gaps, and elongated structures between ore body contours, construct a spatial segmentation domain as a constraint, and generate a three-dimensional model of multiple ore bodies that conforms to geological understanding under this constraint. Specifically: (1) The model has high fitting accuracy and completely eliminates overlapping gap defects. This application constructs an external spatial segmentation domain to forcibly define the precise contact interface of multiple ore bodies, unifies the triangular network topology of the contact surface of adjacent ore bodies, and eliminates the fitting deviation of the coplanar region of ore bodies from the root of the algorithm. Specifically, firstly, based on the overlap judgment tolerance, all sub-segment pairs of the two ore body contours are traversed within the same profile, and the overlapping plane dividing lines are accurately extracted through the double vertical distance judgment condition. This condition requires that the vertical distance between the vertices of the two line segments is less than the tolerance and that at least one vertex falls within the body range of the other line segment. Then, a similarity matrix is ​​established for the overlapping plane dividing lines extracted from adjacent profiles. The maximum cumulative similarity is solved through dynamic programming and the optimal matching pair set is generated by backtracking, so as to achieve the order-preserving matching of cross-profile dividing lines. Then, the matched overlapping dividing lines are connected to generate an external space dividing domain. Finally, when both ends of a certain segment are overlapping dividing domains, the surface of the ore body corresponding to that segment is not generated by dynamic programming mapping, but directly uses the surface patch of the external space dividing domain, so that the adjacent ore bodies share the same surface patch at the contact position. The vertex coordinates, connection order and normal direction of the triangular mesh are completely consistent, so as to achieve complete fit, no overlap and no gap at the contact position of multiple ore bodies, and perfectly solve the fitting distortion problem of traditional modeling. (2) Effectively avoid model interference and self-crossing, and has strong anti-distortion ability. This application constructs an internal spatial segmentation domain by decomposing and extracting the midpoint skeleton of a narrow domain, rigidly constraining the connection path of the narrow ore body contour, and isolating adjacent ore body spaces through an external segmentation domain. Specifically, for a narrow ore body, the overall shape factor is first calculated for macroscopic screening. Constrained triangulation is performed on the contour determined to be narrow—forcing only triangles located at the two ends of the contour to have two sides that fit the contour line, while the remaining triangles have only one side that fits the contour line. Then, the vertical distance from the vertex to the reference edge is calculated using the edge that fits the contour line as the reference edge. Triangles smaller than the narrow width threshold are marked as narrow triangles and connected to form a narrow domain. Based on the topological relationship between the narrow domain and the contour line, it is divided into three types: fully closed, semi-open, and fully open. Fully closed and semi-open narrow domains are decomposed into two independent polylines using the common vertex of the characteristic triangles as the cutoff point. Then, an open line segment connecting frame algorithm is used to construct a triangular network and extract the midpoints of the internal edges and connect them in an orderly manner to form a narrow planar segmentation line. After cross-section matching, the connection is used to generate an internal spatial segmentation domain, forcing the connection path of the narrow ore body contour to extend along the skeleton line.For adjacent gaps between ore bodies, adjacent line segments are screened and overlapping areas are removed based on the gap identification threshold to generate gap plane dividing lines. After cross-section matching, they are connected to generate an external spatial segmentation domain. In the subsequent dynamic programming modeling, a forbidden point set mechanism is introduced. Point pairs that intersect with the spatial segmentation domain after connection are added to the forbidden point set and the cost value is set to infinity, which mathematically forces the blocking of outbound paths. Through the dual mechanism of internal domain constraint direction and external domain isolation space, model interference between adjacent ore bodies is avoided, and the distortion and self-intersection problems of narrow ore body modeling are completely solved, which greatly improves the topological rationality of complex ore body models. (3) Full-process automation adapts to complex geological scenarios. This application realizes the full-process automation of ore body contour overlap, gap, narrow structure automatic identification, plane dividing line automatic extraction, cross-section order-preserving matching, and constraint mapping modeling. Specifically, based on overlap judgment tolerance, elongated shape factor threshold, elongated width threshold, and gap recognition threshold, the system automatically identifies and extracts corresponding overlapping plane segmentation lines, elongated plane segmentation lines, and gap plane segmentation lines. It automatically establishes a similarity matrix by treating the segmentation line sets of adjacent profiles as sequences, solves for the maximum cumulative similarity through dynamic programming, and backtracks to generate the optimal matching pair set, achieving automatic matching across profiles while maintaining order. Based on the matching results, it automatically constructs internal and external spatial segmentation domains. Using the spatial segmentation domain as a hard constraint, it automatically extracts the shortest connection path and generates a 3D wireframe model through a dynamic programming algorithm with a prohibited point set. The entire modeling process requires no manual correction of various model defects and can adapt to various complex geological conditions such as multi-orebody contact, adjacent gaps, and elongated complex ore bodies, significantly improving modeling efficiency and versatility. Attached Figure Description

[0058] Figure 1 A flowchart illustrating the multi-orebody contour modeling method provided in this application embodiment;

[0059] Figure 2 This is a schematic diagram of a fully enclosed, narrow domain structure provided in an embodiment of this application;

[0060] Figure 3 A schematic diagram of a semi-open, narrow domain structure provided in an embodiment of this application;

[0061] Figure 4 A schematic diagram of a fully open narrow domain structure provided in an embodiment of this application;

[0062] Figure 5 This is a schematic diagram of the internal space partitioning domain structure provided in an embodiment of this application;

[0063] Figure 6 This is a schematic diagram of the external space partitioning domain structure provided in an embodiment of this application;

[0064] Figure 7A flowchart illustrating the multi-orebody contour modeling method based on spatial segmentation domain constraints, provided as an application example of this application;

[0065] Figure 8 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0066] The present application will now be described in further detail with reference to the accompanying drawings and embodiments.

[0067] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application.

[0068] This application provides a method for modeling the contour lines of multiple ore bodies, such as... Figure 1 As shown, the specific steps are as follows:

[0069] Step 110: Obtain multiple ore body contour lines and preset threshold parameters on adjacent profiles. The multiple ore body contour lines are a set of points arranged in counterclockwise order.

[0070] It is understood that adjacent profiles include a first profile P and a second profile Q adjacent to the first profile, where the first profile P and the second profile Q are two adjacent profiles on the exploration line. Multiple orebody outlines are the orebody boundary lines interpreted on each profile, used to characterize the cross-sectional shape of the orebody on the profile. In this embodiment, the orebody outline is a closed polygon, composed of multiple point sets arranged in a counter-clockwise order.

[0071] For example, suppose two adjacent cross sections are respectively and .profile The area containing the outlines of two ore bodies is denoted as a closed polygon. , ;profile The corresponding outline is denoted as , Each contour line is a set of points arranged counterclockwise in the plane:

[0072] ;

[0073] It should be noted that the above example uses two profiles, each containing two ore body outlines. In actual geological exploration, the number of adjacent profiles and the number of ore body outlines on each profile can be determined based on actual exploration data. This application does not impose specific limitations on this. The purpose of arranging the point set in a counterclockwise order is to ensure that the normal direction of each outline is consistent, conforming to the right-hand rule. This ensures that the normal orientation of all triangular faces in the subsequently constructed triangular mesh model is uniform, avoiding rendering errors or volume calculation deviations caused by reversed normals in some faces. At the same time, the orderly arrangement ensures that the adjacent vertices of the outlines are clear, providing a basis for subsequent traversal of sub-segment pairs.

[0074] Here, the preset threshold parameters are used to control the recognition sensitivity of various geometric features in subsequent algorithms, including overlap judgment tolerance, gap recognition threshold, narrow width threshold, and narrow shape factor threshold.

[0075] For example, the following threshold parameters are defined: (1) Overlap determination tolerance (1) Distance threshold for determining whether two sides are collinear; (2) Gap recognition threshold : Determine whether the maximum distance between two contours is a gap; (3) Narrow width threshold : Areas smaller than this width are considered as local narrow regions; (4) Narrow shape factor threshold Used for preliminary determination of the overall narrow length.

[0076] Here, the overall shape factor of the ore body outline is defined as f = 4πA / L², where A is the area enclosed by the outline and L is the perimeter of the outline. When f < λthin, the ore body outline is determined to be elongated. It is understood that the specific values ​​of the above threshold parameters can be set by those skilled in the art according to actual geological conditions and modeling requirements, and this application embodiment does not specifically limit them.

[0077] Here, step 110 is the input stage of this application embodiment, which obtains the basic data required for subsequent modeling and provides a standardized input format and adjustable control parameters for subsequent steps.

[0078] Step 120: Based on preset threshold parameters and multiple ore body contour lines corresponding to each profile in adjacent profiles, determine the set of planar dividing lines for each profile in adjacent profiles; wherein, the preset threshold parameters include: overlap judgment tolerance, elongated shape factor threshold, elongated width threshold, and gap identification threshold; the set of planar dividing lines includes: overlapping planar dividing lines, elongated planar dividing lines, and gap planar dividing lines; the set of planar dividing lines includes the first set of planar dividing lines for the first profile and the second set of planar dividing lines for the second profile adjacent to the first profile.

[0079] Here, for each adjacent profile, overlapping planar segmentation lines between multiple ore body contour lines within that profile are extracted based on overlap determination tolerance; narrow planar segmentation lines between each ore body contour line within that profile are extracted based on narrow shape factor threshold and narrow width threshold; and gap planar segmentation lines between multiple ore body contour lines within that profile are extracted based on gap identification threshold. The extraction order of these three planar segmentation lines is not limited in this embodiment and can be flexibly adjusted according to actual needs.

[0080] Understandably, overlapping plane dividing lines are used to characterize the boundary segments of coplanar contact between the contour lines of two ore bodies within the same profile. When extracting overlapping plane dividing lines, all sub-segment pairs of multiple ore body contours within the profile are traversed. A preset overlap judgment tolerance is used as the judgment threshold. The existence of local overlap is determined based on the vertical distance relationship between each sub-segment pair. Overlapping segments are extracted and continuous overlapping segments are merged, ultimately forming an ordered point sequence. By extracting overlapping plane dividing lines, the shared boundary between the two ore body contour lines is independently separated from the original contours, providing a geometric basis for subsequently constructing the contact interface between ore bodies.

[0081] Understandably, elongated planar dividing lines are used to characterize the central skeleton path of elongated regions within the ore body contour. When extracting these lines, the overall shape factor is first used to determine if the ore body contour exhibits elongated characteristics. If so, constrained triangulation is performed within the contour. Narrow triangles are marked according to a width threshold and connected to form elongated domains. These domains are then classified into three types based on topological relationships: fully enclosed, semi-open, and fully open, and preprocessed. Finally, a triangulation network is constructed from the preprocessed elongated domains, and the midpoints of the internal edges are extracted and connected to form skeleton lines. The extraction of elongated planar dividing lines provides internal guidance for constraining the orientation of elongated ore bodies and preventing twisting self-intersections during subsequent cross-section connections.

[0082] Understandably, the fracture plane segmentation line is used to characterize the center line of the gap between two non-overlapping but very close orebody contours within the same profile. When extracting the fracture plane segmentation line, all boundary sub-segment pairs of multiple orebody contours within the profile are traversed. Neighboring segments with a nearest point distance less than the fracture identification threshold are selected, and areas already determined to be overlapping are removed to obtain fracture boundary segments. An open segment wireframe algorithm is used to construct a triangular mesh for these fracture boundary segments. The midpoints of the internal non-boundary edges are extracted and sequentially connected according to the fracture topology extension direction to form a skeleton line. The extraction of the fracture plane segmentation line provides geometric constraints for subsequent construction of isolation interfaces between orebody segments and prevention of model over-boundary interference.

[0083] It should be noted that the specific algorithm details involved in the above three methods for extracting planar segmentation lines, including line segment overlap determination, constrained triangulation, narrow domain classification and preprocessing, and open line segment wireframe algorithm, will be elaborated in subsequent steps and will not be repeated here.

[0084] Here, after extracting the three types of planar segmentation lines, the overlapping planar segmentation lines, the elongated planar segmentation lines, and the gap planar segmentation lines together constitute a complete set of section segmentation lines for this profile. Among them, the overlapping planar segmentation lines and the elongated planar segmentation lines are uniformly arranged in a counterclockwise spiral order according to the ore body contour, while the gap planar segmentation lines are arranged in an orderly manner according to the topological extension direction of the elongated domain. This provides regular input data for subsequent cross-section ordered structure matching and spatial segmentation domain construction.

[0085] In step 120, the main task is data preprocessing, which involves extracting planar dividing lines (overlapping dividing lines, fracture skeleton lines, and elongated skeleton lines) from the contours of each group of ore bodies. By extracting overlapping planar dividing lines, elongated planar dividing lines, and fracture planar dividing lines respectively and forming a set of planar dividing lines for each profile, the topological relationships (overlap, fractures) between ore bodies within the same profile and the morphological characteristics (elongated) of individual ore bodies are uniformly transformed into standardized line descriptions, providing structured input data for subsequent cross-profile matching and spatial constraint domain construction.

[0086] Step 130: Take the first set of plane segmentation lines and the second set of plane segmentation lines as the first plane segmentation sequence and the second plane segmentation sequence, respectively, and establish a similarity matrix between the first plane segmentation sequence and the second plane segmentation sequence; solve the maximum cumulative similarity function constructed by the similarity matrix based on dynamic programming, and generate the optimal matching pair set based on the path backtracking corresponding to the maximum cumulative similarity; based on the optimal matching pair set, construct the internal spatial segmentation domain between the narrow plane segmentation lines of two adjacent profiles and the external spatial segmentation domain between the overlapping plane segmentation lines or gap plane segmentation lines of two adjacent profiles.

[0087] Understandably, in this embodiment, after the processing in step 120, respective sets of section dividing lines are extracted from the first section P and the second section Q. Each set contains several planar dividing lines, which may be overlapping planar dividing lines, narrow planar dividing lines, or gap planar dividing lines. In order to construct a constraint surface in three-dimensional space, it is necessary to establish a one-to-one correspondence between the dividing lines on the first section P and the dividing lines on the second section Q according to their types. This process is called order-preserving structure matching.

[0088] Specifically, the set of planar dividing lines of the first profile P is taken as the first planar dividing sequence A, and the set of planar dividing lines of the second profile Q is taken as the second planar dividing sequence B. An optimal correspondence between A and B is established, allowing skipping some elements without changing the original order of the two sequences. First, a similarity matrix is ​​established: for each pair of dividing lines in the two sequences, three dimensions are considered comprehensively: length similarity (measuring the closeness of the total length), positional closeness (measuring whether the spatial positions on the profile are close), and directional consistency (measuring whether the extension directions are consistent). A weighted sum is obtained to obtain the comprehensive similarity, and all combinations are traversed to construct a complete similarity matrix. Then, dynamic programming is used to solve for the maximum cumulative similarity—traversing the similarity matrix, selecting the optimal path from three options at each step: pairing the two dividing lines (diagonal direction), skipping the current element of the first sequence (upward direction), or skipping the current element of the second sequence (leftward direction). After traversal, the optimal path is traced in reverse order from the endpoint, extracting all matching positions along the path to generate the optimal matching pair set. The indices of the matching pairs in both sequences are monotonically increasing to maintain the original order. It should be understood that the core of order-preserving matching is to not change the original arrangement order of the dividing lines on their respective cross sections, because this order reflects the spatial topological relationship of the ore body. Changing the order will lead to incorrect cross connections in the 3D model.

[0089] Here, spatial segmentation domains will be constructed. Specifically, based on the set of optimal matching pairs, internal and external spatial segmentation domains will be constructed respectively. If the ore bodies corresponding to two adjacent cross-sections are both elongated ore bodies, the matching pairs corresponding to the elongated planar dividing lines in the optimal matching pairs will be taken as internal matching pairs. The open segment wireframe algorithm will be used to construct triangular mesh surfaces from the two elongated planar dividing lines corresponding to each matching pair. All surfaces will be combined to form the internal spatial segmentation domain, which will be used to constrain the connection direction of the elongated ore bodies and prevent distortion and self-intersection. If there are overlapping or gapped planar dividing lines on two adjacent cross-sections, the matching pairs corresponding to them in the optimal matching pairs will be taken as external matching pairs. The two dividing lines corresponding to each matching pair will be connected to generate triangular mesh surfaces. All surfaces will be combined to form the external spatial segmentation domain, which will define the contact interface (overlapping scenario) or isolation interface (gap scenario) between the two ore bodies to prevent model penetration or interference.

[0090] Here, step 130 establishes the correspondence between the dividing lines of adjacent sections through order-preserving structure matching, and then constructs the internal and external spatial dividing domains respectively, transforming the geometric feature lines on the two-dimensional section into hard constraint surfaces in three-dimensional space, providing a mandatory spatial constraint boundary for subsequent constraint elastic synchronous mapping modeling.

[0091] Step 140: Use the endpoints of each element in the optimal matching pair set as forced anchor points on the ore body contours of the first and second profiles. Divide each ore body contour into multiple one-to-one corresponding segments based on the forced anchor points. For each segment, using the internal and external spatial partitioning domains as hard constraints, extract the shortest connection path between the starting and ending anchor points of the segment that does not intersect with the internal and external spatial partitioning domains based on a dynamic programming algorithm with a forbidden point set. Generate a connection triangular network for each ore body based on each shortest connection path. Construct a multi-ore body 3D wireframe model based on each connection triangular network.

[0092] In this embodiment, after the processing in step 130, the internal spatial segmentation domain and the external spatial segmentation domain have been constructed. The core task of step 140 is to use these spatial segmentation domains as hard constraints to generate connecting triangular meshes between the ore body contours of adjacent profiles, thereby constructing a multi-ore body three-dimensional wireframe model.

[0093] Specifically, the endpoints of each element in the optimal matching pair set obtained in step 130 are used as forced anchor points. Since each element in the optimal matching pair set represents the correspondence between a dividing line of the first profile and a dividing line of the second profile, the endpoints of these dividing lines are the corresponding feature points of the ore body contours on the two profiles. Using these endpoints as forced anchor points means that the subsequently generated triangular mesh must pass through these points, thus ensuring that the endpoints of the dividing lines are precisely aligned in three-dimensional space. For example, the two endpoints of an overlapping dividing line correspond to the start and end points of the coplanar contact area of ​​two ore bodies, respectively. Using these as forced anchor points locks the boundary position of this coplanar area in three-dimensional space, ensuring that the contact positions of the two ore bodies in this area are strictly consistent.

[0094] After determining the mandatory anchor points, the original complete ore body contour is divided into multiple corresponding segments according to the arrangement order of the anchor points on the ore body contour. The start and end points of each segment are mandatory anchor points, and the contour point sequences within the corresponding segments on the first and second profiles constitute the point set to be matched. The purpose of segmenting the contour is to decompose the large-scale contour matching problem into multiple independent sub-problems, and to perform matching separately within each segment, thereby reducing computational complexity while ensuring that the anchor point positions are forcibly traversed.

[0095] For each segment, a dynamic programming algorithm with a prohibited point set is used to extract the shortest connection path between the starting and ending anchor points of that segment. The prohibited point set is defined according to the spatial partitioning domain: if a triangular facet generated by connecting a point on the first profile and a point on the second profile intersects any internal or external spatial partitioning domain, then that point pair is added to the prohibited point set. During the dynamic programming recursion, the cumulative cost of the current position is determined by the minimum of the distance cost between points plus the cumulative cost of migrating from the previous position. Specifically, the current position can be obtained by migrating in three directions: diagonally, which means pairing the current point on the first profile with the current point on the second profile; upward, which means skipping the current point on the first profile; and leftward, which means skipping the current point on the second profile. However, migration in any of these directions must satisfy the legal constraint that the newly generated triangular facet during the migration process does not intersect any spatial partitioning domain, and the intermediate point is not within the prohibited point set. If a migration in a certain direction does not meet the legal constraints, then the cost of that direction is set to infinity, making it impossible for it to be the optimal choice. Through this method, dynamic programming ensures that the extracted connection path is geometrically shortest while strictly avoiding all internal and external spatial partitions.

[0096] It is worth emphasizing that, in the special case where both ends of a segment are overlapping partitioned domains, the surface of the ore body corresponding to that segment is not generated through the aforementioned dynamic programming mapping. Instead, the surface patch of the external spatial partitioned domain corresponding to the overlapping partitioned domain is directly used. This is because an overlapping partitioned domain indicates that two ore bodies are in coplanar contact in that area. Directly using the pre-constructed external spatial partitioned domain surface patch ensures that the vertex coordinates, connection order, and normal direction of the triangulation of adjacent ore bodies at that location are completely consistent, thereby eliminating gaps or penetrations that may occur in the coplanar region from the root. This special treatment is one of the important features that distinguishes this embodiment from the traditional independent modeling method.

[0097] After extracting the shortest connection paths for each section, the point pairs on all paths are connected sequentially to generate a triangular network connecting each ore body between adjacent profiles, thereby constructing a complete three-dimensional wireframe model of multiple ore bodies.

[0098] In step 140, the precise alignment of the endpoints of the cross-section dividing line is ensured by the forced anchor point, the geometric optimality of the connection path is ensured by the dynamic programming algorithm with the forbidden point set, and the seamless fit of the coplanar contact area is ensured by directly using the external spatial dividing domain surface patch for the overlapping section. Finally, the accurate construction of the three-dimensional wireframe model of the multi-ore body is achieved.

[0099] Thus, given the existence of multiple ore body contours on adjacent exploration profiles, this application provides a collaborative modeling method for multiple ore body contours, which can automatically identify overlaps, gaps, and elongated structures between ore body contours, construct a spatial segmentation domain as a constraint, and generate a three-dimensional model of multiple ore bodies that conforms to geological understanding under this constraint. Specifically: (1) The model has high fitting accuracy and completely eliminates overlapping gap defects. This application constructs an external spatial segmentation domain to forcibly define the precise contact interface of multiple ore bodies, unifies the triangular network topology of the contact surface of adjacent ore bodies, and eliminates the fitting deviation of the coplanar region of ore bodies from the root of the algorithm. Specifically, firstly, based on the overlap judgment tolerance, all sub-segment pairs of the two ore body contours are traversed within the same profile, and the overlapping plane dividing lines are accurately extracted through the double vertical distance judgment condition. This condition requires that the vertical distance between the vertices of the two line segments is less than the tolerance and that at least one vertex falls within the body range of the other line segment. Then, a similarity matrix is ​​established for the overlapping plane dividing lines extracted from adjacent profiles. The maximum cumulative similarity is solved through dynamic programming and the optimal matching pair set is generated by backtracking, so as to achieve the order-preserving matching of cross-profile dividing lines. Then, the matched overlapping dividing lines are connected to generate an external space dividing domain. Finally, when both ends of a certain segment are overlapping dividing domains, the surface of the ore body corresponding to that segment is not generated by dynamic programming mapping, but directly uses the surface patch of the external space dividing domain, so that the adjacent ore bodies share the same surface patch at the contact position. The vertex coordinates, connection order and normal direction of the triangular mesh are completely consistent, so as to achieve complete fit, no overlap and no gap at the contact position of multiple ore bodies, and perfectly solve the fitting distortion problem of traditional modeling. (2) Effectively avoid model interference and self-crossing, and has strong anti-distortion ability. This application constructs an internal spatial segmentation domain by decomposing and extracting the midpoint skeleton of a narrow domain, rigidly constraining the connection path of the narrow ore body contour, and isolating adjacent ore body spaces through an external segmentation domain. Specifically, for a narrow ore body, the overall shape factor is first calculated for macroscopic screening. Constrained triangulation is performed on the contour determined to be narrow—forcing only triangles located at the two ends of the contour to have two sides that fit the contour line, while the remaining triangles have only one side that fits the contour line. Then, the vertical distance from the vertex to the reference edge is calculated using the edge that fits the contour line as the reference edge. Triangles smaller than the narrow width threshold are marked as narrow triangles and connected to form a narrow domain. Based on the topological relationship between the narrow domain and the contour line, it is divided into three types: fully closed, semi-open, and fully open. Fully closed and semi-open narrow domains are decomposed into two independent polylines using the common vertex of the characteristic triangles as the cutoff point. Then, an open line segment connecting frame algorithm is used to construct a triangular network and extract the midpoints of the internal edges and connect them in an orderly manner to form a narrow planar segmentation line. After cross-section matching, the connection is used to generate an internal spatial segmentation domain, forcing the connection path of the narrow ore body contour to extend along the skeleton line.For adjacent gaps between ore bodies, adjacent line segments are screened and overlapping areas are removed based on the gap identification threshold to generate gap plane dividing lines. After cross-section matching, they are connected to generate an external spatial segmentation domain. In the subsequent dynamic programming modeling, a forbidden point set mechanism is introduced. Point pairs that intersect with the spatial segmentation domain after connection are added to the forbidden point set and the cost value is set to infinity, which mathematically forces the blocking of outbound paths. Through the dual mechanism of internal domain constraint direction and external domain isolation space, model interference between adjacent ore bodies is avoided, and the distortion and self-intersection problems of narrow ore body modeling are completely solved, which greatly improves the topological rationality of complex ore body models. (3) Full-process automation adapts to complex geological scenarios. This application realizes the full-process automation of ore body contour overlap, gap, narrow structure automatic identification, plane dividing line automatic extraction, cross-section order-preserving matching, and constraint mapping modeling. Specifically, based on overlap judgment tolerance, elongated shape factor threshold, elongated width threshold, and gap recognition threshold, the system automatically identifies and extracts corresponding overlapping plane segmentation lines, elongated plane segmentation lines, and gap plane segmentation lines. It automatically establishes a similarity matrix by treating the segmentation line sets of adjacent profiles as sequences, solves for the maximum cumulative similarity through dynamic programming, and backtracks to generate the optimal matching pair set, achieving automatic matching across profiles while maintaining order. Based on the matching results, it automatically constructs internal and external spatial segmentation domains. Using the spatial segmentation domain as a hard constraint, it automatically extracts the shortest connection path and generates a 3D wireframe model through a dynamic programming algorithm with a prohibited point set. The entire modeling process requires no manual correction of various model defects and can adapt to various complex geological conditions such as multi-orebody contact, adjacent gaps, and elongated complex ore bodies, significantly improving modeling efficiency and versatility.

[0100] In some embodiments, based on a preset threshold parameter and multiple ore body contour lines corresponding to each profile in adjacent profiles, a set of planar dividing lines for each profile in adjacent profiles is determined, including:

[0101] Based on the overlap determination tolerance, the overlapping plane dividing line between multiple ore body outlines in each of the adjacent profiles is determined.

[0102] Based on the narrow shape factor threshold and the narrow width threshold, the narrow plane dividing line between multiple ore body outlines in each profile of adjacent profiles is determined.

[0103] Based on the gap identification threshold, the gap plane dividing line between multiple ore body contours in each of the adjacent profiles is determined;

[0104] A set of plane segmentation lines is generated based on overlapping plane segmentation lines, narrow plane segmentation lines, and gap plane segmentation lines.

[0105] Among them, based on the overlap determination tolerance, the overlapping plane dividing line between multiple ore body contours in each of adjacent profiles is determined, including:

[0106] For each of the adjacent profiles, iterate through all sub-line pairs of the contours of multiple ore bodies within the profile.

[0107] Extract overlapping sub-segment pairs from all sub-segments that satisfy a first set of conditions and a second set of conditions. The first set of conditions is that the first perpendicular distance between the two vertices of the first sub-segment and the second sub-segment or its extension in the sub-segment pair is less than the overlap judgment tolerance, and the second perpendicular distance between the two vertices of the second sub-segment and the first sub-segment or its extension is less than the overlap judgment tolerance. The second set of conditions is that at least one vertex of the first sub-segment and the second sub-segment has a perpendicular distance between it and the first sub-segment or its extension that is less than the overlap judgment tolerance. Extract overlapping segments from the overlapping sub-segment pairs and merge consecutive overlapping segments. Based on each consecutive overlapping segment, generate a set of overlapping plane dividing lines for the cross-section. Based on the set of overlapping plane dividing lines for each cross-section, generate a set of overlapping plane dividing lines.

[0108] Among them, based on the gap identification threshold, the plane dividing line of the gap between multiple ore body contours in each profile of adjacent profiles is determined, including:

[0109] For each profile in an adjacent profile, traverse all boundary sub-segment pairs of multiple ore body contours within the profile, and extract the nearest neighbor segments between all boundary sub-segment pairs whose distance to the nearest point is less than the gap identification threshold.

[0110] Remove areas where adjacent line segments have been determined to overlap, and generate gap boundary segments;

[0111] Based on the narrow domain skeleton extraction algorithm, a triangular network is constructed for the gap boundary segment; based on the gap topological extension direction, the midpoints of all internal non-boundary edges in the triangular network are sequentially connected to generate the gap plane segmentation line.

[0112] In this embodiment, determining the set of planar dividing lines includes extracting three types of dividing lines. It is understood that the extraction order of the three types of dividing lines can be flexibly adjusted according to actual needs, and this embodiment does not impose specific limitations on this.

[0113] For example, the extraction of overlapping parts between contours is specifically as follows: for two ore body contours within the same profile... The outline of the first ore body and The second ore body outline is obtained by traversing all sub-segment pairs of the two outlines, i.e., line segments. With line segment Accurately determine the overlapping relationship of line segments based on distance constraints, and adopt a preset overlap determination tolerance. This serves as a threshold for determination. If a pair of line segments simultaneously meets both of the following conditions, then the two line segments are determined to have local overlap:

[0114] The first condition is: line segment From the two vertices to the line segment or the perpendicular distance of its extension, and the line segment From the two vertices to the line segment The perpendicular distance between the line or its extension is less than the overlap determination tolerance. ;

[0115] The second condition is: , There exists at least one line segment whose vertex has a perpendicular distance from another original line segment (excluding its extension) less than 1 / 2. For line segment pairs determined to be overlapping, the overlapping segments are further extracted, and all consecutive overlapping segments are merged to form an ordered point sequence, ultimately yielding the overlapping plane dividing line, denoted as... .

[0116] It is understandable that both the first and second conditions must be satisfied simultaneously, and there is an "AND" logical relationship between them. The first condition ensures that the two sub-segments are collinear or nearly collinear, while the second condition ensures that the two sub-segments actually overlap in space. Through this dual determination mechanism, truly coplanar contact boundary segments can be accurately identified, while non-overlapping segment segments with the same direction but spatially offset are excluded.

[0117] It should be noted that the specific method for extracting the elongated plane segmentation line will be described in detail in subsequent embodiments. The core idea is as follows: First, determine whether the ore body outline is elongated by the overall shape factor. If it is determined to be elongated, perform constrained triangulation on the outline. Identify narrow triangles according to the elongated width threshold and connect them to form an elongated domain. Then, divide the elongated domain into three types: fully enclosed, semi-open, and fully open, and perform preprocessing. Finally, extract the skeleton line from the preprocessed elongated domain as the elongated plane segmentation line.

[0118] For example, the content of gap region identification and segmentation line generation is as follows:

[0119] For two sets of ore body outlines that do not overlap and are adjacent to each other within the same profile and Iterate through all boundary sub-segment pairs and filter out the closest points whose distance is less than the gap detection threshold. The adjacent line segments are identified, and overlapping areas are removed to obtain pure fracture boundary segments. The effective fracture consists of two boundary segments belonging to two different ore bodies and isolated from each other, with a topological structure consistent with the fully open narrow domain.

[0120] For the bi-segment independent polyline structure of the narrow domain of the gap, an open segment wireframe algorithm is used to construct a triangular mesh. The narrow domain skeleton extraction algorithm is reused: the midpoints of all internal non-boundary edges of the triangular mesh are extracted as skeleton feature nodes. These nodes are then sequentially connected along the gap's topological extension direction to generate a smooth skeleton path that fits the geometric center of the gap. The resulting ordered chain of nodes is the gap plane dividing line, denoted as […]. .

[0121] Understandably, since the gap region consists of two isolated boundary segments belonging to two different ore bodies, its geometric topology is completely consistent with that of a fully open narrow domain. Therefore, the narrow domain skeleton extraction algorithm can be directly reused without designing a separate skeleton extraction process for gaps, thus reducing the implementation complexity of the algorithm.

[0122] In some embodiments, based on a narrow shape factor threshold and a narrow width threshold, determining narrow planar dividing lines between multiple ore body contours in each of adjacent profiles includes:

[0123] For each orebody contour line of multiple orebody contour lines in each adjacent profile, the overall shape factor of the orebody contour line is obtained. If the overall shape factor is less than the elongated shape factor threshold, constrained triangulation is performed on the interior of the orebody contour line to generate multiple constrained triangulation triangles. Among these multiple constrained triangulation triangles, the positional constraint condition is: among the multiple constrained triangulation triangles, only the triangulation triangles located at both ends of the contour line have two sides that fit the orebody contour line, and all other internal triangles have only one side that fits the orebody contour line. For each constrained triangulation triangle, the vertex opposite the constrained triangulation triangle is calculated using the side of the constrained triangulation triangle that fits the orebody contour line as the reference side. The third perpendicular distance from a point to the reference edge is used to determine if the triangle is a narrow triangle if the third perpendicular distance is less than the narrow width threshold. Connecting all adjacent narrow triangles forms a narrow domain of the ore body. Based on the topological relationship between the narrow domain and the ore body outline, the type of the narrow domain is determined. The types of narrow domains include: fully enclosed narrow domains, semi-open narrow domains, and fully open narrow domains. A fully enclosed narrow domain contains two characteristic triangles, with two sides of each triangle fitting the ore body outline. A semi-open narrow domain contains a unique characteristic triangle, with two sides fitting the ore body outline. A fully open narrow domain does not contain any characteristic triangles with two sides fitting the ore body outline.

[0124] Based on the type of the narrow domain of the ore body, a target narrow domain is generated; for the target narrow domain, a triangular mesh is constructed; the midpoints of all internal edges in the triangular mesh that do not fit with the outline of the ore body are extracted, and the midpoints are connected in an orderly manner according to the topological extension direction to construct the narrow plane dividing line of the ore body outline.

[0125] Based on the narrow plane segmentation lines of each ore body outline, the narrow plane segmentation lines of the profile are generated.

[0126] Based on the narrow plane dividing lines of each profile, generate narrow plane dividing lines.

[0127] In some embodiments, generating a target narrow domain based on the type of the ore body narrow domain includes:

[0128] If the type of the narrow domain of the ore body is a fully enclosed narrow domain or a semi-open narrow domain, then the narrow domain of the ore body is preprocessed to generate the target narrow domain;

[0129] If the type of the narrow domain of the ore body is a fully open narrow domain, then the fully open narrow domain is the target narrow domain;

[0130] If the narrow ore body is a fully enclosed or semi-open narrow ore body, then the narrow ore body is preprocessed to generate the target narrow ore body, including:

[0131] If the type of the narrow domain of the ore body is a fully enclosed narrow domain, then the common vertex of the two characteristic triangles with two sides that fit the outline of the ore body within the fully enclosed narrow domain is used as the cut-off point to decompose the closed narrow domain into two independent polylines that are not connected to each other; based on each independent polyline that is not connected to each other, the preprocessed narrow domain is generated.

[0132] If the type of the ore body narrow domain is a semi-open narrow domain, then the common vertex of the characteristic triangle whose only two sides in the semi-open narrow domain region conform to the topological features of the ore body outline is used as the cutoff point. The semi-open narrow domain is then decomposed into two independent polylines that are not connected to each other. Based on the two independent polylines that are not connected to each other, the pre-processed narrow domain is generated.

[0133] For example, the specific details of the technical solution for narrow region recognition and skeleton line extraction are as follows:

[0134] First, calculate the overall shape factor of the single ore body outline. ( The area enclosed by the outline. (where is the perimeter of the outline) The overall outline of the ore body is determined to be elongated. To accurately identify the elongated structure of local parts of the ore body, constrained triangulation is performed on the closed ore body outline. The triangulation satisfies the following topological rules: only the triangulation triangles located at both ends of the outline have two sides that fit the ore body outline, all other internal triangles have only one side that fits the ore body outline, and the remaining two sides of the triangles are all located within the ore body outline, ensuring that the triangulation structure fits the topological features of the ore body boundary. After completing the triangulation, the elongation attribute of each triangulation triangle is determined: for a single triangle, the side that fits the ore body outline is extracted as the reference side, and the vertical distance from the vertex of the triangle opposite the reference side to the reference side is calculated. This vertical distance is used as the basis for determining the local width of the triangle. If this vertical distance is less than a preset elongation width threshold... If the triangle is narrow, it is marked as such. Connect all adjacent narrow triangles to form a complete and continuous narrow domain of the ore body. Based on the topological relationship between the narrow domain and the ore body outline, the narrow domain is divided into three types: fully enclosed, semi-open, and fully open, and preprocessing is performed accordingly.

[0135] Fully Enclosed Narrow Domain: A fully enclosed narrow domain is a continuous, narrow, closed region corresponding to the entire ore body outline. Within this region, there are two characteristic triangles. These triangles possess the topological characteristic that two sides conform to the ore body outline. The common vertex of the two outline sides of this characteristic triangle is located, and this vertex is used as the cutoff point to break the completely closed narrow domain. This deconstructs the originally closed narrow domain into two independent, unconnected polylines, completing the pre-processing of the fully enclosed narrow domain.

[0136] For example, such as Figure 2 As shown in the figure, this is a schematic diagram of the structure of a fully enclosed narrow domain. The black double lines represent the outer contour boundary of the ore body, and the light gray diagonal lines inside are the internal grid edges of the constrained triangulation. The ore body as a whole has a slender closed loop shape with its head and tail connected. The entire domain is a closed narrow region formed by the connection of narrow triangles. There is a feature triangle with two sides attached to the outer boundary at each end of the contour, which constitutes a fully enclosed narrow domain. This type of narrow domain has no natural segmentation at the head and tail of the closed loop. It is necessary to locate the common vertex of the feature triangle as the cutoff point and disassemble the closed loop to obtain two independent polylines.

[0137] Semi-open narrow region: A semi-open narrow region is a local narrow, non-closed area at the end of the ore body outline. Within this region, there exists a unique characteristic triangle, which also has the topological feature that two sides fit the ore body outline. By locking the common vertex of the two outline sides of this characteristic triangle, and using this as the cutoff point, the semi-open narrow region is broken apart and disassembled, ultimately splitting the semi-open narrow region into two independent, unconnected polyline structures.

[0138] For example, such as Figure 3 As shown, Figure 3This is a schematic diagram of a semi-open, narrow domain. The black line segment in the diagram represents the straight outer contour of the main body of the ore body. The large, raised area on the right side of the image represents the semi-open, narrow domain formed by the extension of the end. The narrow domain is connected to the main body of the ore body on only one side and extends freely on the other side. There is only one set of feature triangles with double contour edges, which is a typical semi-open, narrow topological structure. It is necessary to lock the common vertex of the two contour edges of the feature triangle to complete the truncation and splitting.

[0139] Fully open narrow domain: The fully open narrow domain is two isolated and unconnected narrow line segments distributed in the middle area of ​​the ore body outline. It is itself a two independent and unconnected polyline structure, which does not require disassembly preprocessing and can be directly involved in the subsequent triangular network reconstruction and skeleton extraction process.

[0140] For example, such as Figure 4 As shown, Figure 4 This is a schematic diagram of the structure of a fully open, narrow domain. The black double lines in the middle of the diagram represent the natural narrow channel in the middle of the ore body, while the left and right sides are the wide main body of the ore body. The narrow channel in the middle is isolated by the wide parts on both sides, naturally forming two segments of narrow polysegment lines that are not connected to each other. The narrow region does not have closed loops or single-end closed structures, and there are no characteristic triangles that need to be truncated, so no additional deconstruction preprocessing is required.

[0141] After unified preprocessing of the three types of narrow domains, all narrow domains are transformed into two independent, disconnected polyline structures. For the decomposed bi-segment polylines, an open segment wireframe algorithm is used to construct a triangulated network. Based on the reconstructed triangulated network, the edge midpoint concatenation method is used to extract the central skeleton of the narrow domain. The complete algorithm flow is as follows:

[0142] Step 1: Reconstruction of Triangulated Networks in Narrow Domains. For the two independent polylines formed after preprocessing, an open segment wireframe algorithm is used to construct a triangulated network.

[0143] Step 2: Extracting the midpoint of the skeleton feature. Traverse all the segmented triangles in the reconstructed triangular mesh, filter out all internal edges that do not conform to the ore body outline, and calculate the coordinates of the center midpoint of each internal edge.

[0144] Step 3: Ordered Serialization of Skeleton Lines. Based on the topological extension direction of the narrow domain, all extracted internal edge midpoints are arranged in an ordered and continuous sequence to form a smooth and continuous skeleton path that fits the geometric center of the narrow domain. The final continuous and ordered chain of nodes is the narrow planar dividing line, denoted as... The dividing line perfectly matches the center trajectory of the narrow domain and strictly adheres to the mandatory endpoint constraints, making it suitable for three types of narrow domain scenarios: fully enclosed, semi-open, and fully open.

[0145] In some embodiments, the similarity matrix is:

[0146] ;

[0147] in, For the i-th dividing line a in the first planar segmentation sequence i With the j-th dividing line in the second plane segmentation sequence Similarity between them For length similarity, For location proximity, The dot product of unit direction vectors. These are the weighting coefficients, This is the sequence of first plane dividing lines. This is the sequence of second-plane dividing lines;

[0148] The maximum cumulative similarity function is:

[0149] ;

[0150] in, From the starting position to the current position The maximum cumulative similarity value, The i-th dividing line in the first planar segmentation sequence With the j-th dividing line in the second plane segmentation sequence The similarity between them.

[0151] For example, the order-preserving structure matching method includes:

[0152] Given a sequence of planar dividing lines of two corresponding ore bodies and An optimal, order-preserving correspondence needs to be established between the two, allowing some elements to be skipped without changing their order.

[0153] Define the similarity matrix ,in,

[0154] ;

[0155] In the formula For length similarity (using the smaller value of the ratio), The proximity is a Gaussian function based on the spatial distance between the midpoints. For directional consistency (dot product of unit direction vectors). These are the weighting coefficients.

[0156] Here, dynamic programming is used to solve for the maximum cumulative similarity:

[0157] ;

[0158] Here, backtracking yields the set of optimal matching pairs. The sequence is monotonically increasing; this method is called order-preserving structure matching.

[0159] In some embodiments, based on the optimal matching pair set, an internal spatial segmentation domain is constructed between the elongated planar segmentation lines of two adjacent sections and an external spatial segmentation domain is constructed between the overlapping planar segmentation lines or gap planar segmentation lines of two adjacent sections, including:

[0160] If the ore bodies corresponding to two adjacent profiles are both narrow and elongated ore bodies, then the matching pairs in the optimal matching pair set that correspond to the narrow and elongated plane dividing lines are taken as internal matching pairs. For each internal matching pair, the open line segment connecting frame algorithm is used to construct the two narrow and elongated plane dividing lines corresponding to the matching pair into a triangular mesh surface. Based on each triangular mesh surface, an internal spatial segmentation domain is constructed.

[0161] If there are overlapping plane dividing lines or gap plane dividing lines on two adjacent cross sections, then the matching pairs in the optimal matching pair set that correspond to the overlapping plane dividing lines or gap plane dividing lines are taken as external matching pairs; for each external matching pair, the two dividing lines corresponding to the matching pair are connected to generate a triangular mesh surface; based on each triangular mesh surface, an external spatial segmentation domain is constructed.

[0162] For example, (1) Construction of internal space segmentation domain (for narrow and elongated scenarios):

[0163] like and All are elongated ore bodies. Order-preserving structural matching is performed on their elongated planar segmentation sequence to obtain matching pairs. Each matching pair... and An open segment wireframe algorithm is used to construct triangular mesh surfaces. All combinations of triangular mesh surfaces constitute the internal spatial segmentation domain of the ore body. This is used to constrain the connection direction of the narrow and elongated parts of the ore body.

[0164] For example, such as Figure 5 As shown in the figure, the gray filled surfaces on the left and right sides are the cross-sectional entities of the narrow ore bodies on the upper and lower sections P and Q, respectively. The black solid lines are the narrow plane dividing lines extracted from the two sections of narrow ore bodies. Multiple light-colored outlines are the complete boundary outlines of the upper and lower ore bodies. The dividing lines are connected by an open line segment frame algorithm to generate continuous triangular mesh surfaces. The triangular mesh surfaces are combined to form the internal spatial division domain Dint that runs through the two layers of ore bodies. This surface extends along the central skeleton of the narrow ore body, accurately constraining the spatial connection direction between the layers of the narrow ore body, and avoiding mesh compression and distortion of the interlayer model.

[0165] (2) Construction of external space segmentation domain (for overlapping / gap scenarios):

[0166] If the cross-section and The surface contains planar dividing lines caused by overlap or gaps. These dividing line sequences are matched in a sequence-preserving manner, and the matched dividing lines are connected to generate a triangular mesh surface, which constitutes the external spatial division domain of the ore body. This segmentation domain directly defines the contact or isolation interface between the two ore bodies, and is a constraint surface that must be strictly controlled in subsequent modeling.

[0167] For example, such as Figure 6 As shown in the figure, the gray filled surfaces on the left and right correspond to the ore body sections with overlapping / gap defects on the upper and lower sections P and Q. The black outline is the boundary of the ore body. The planar dividing lines on the two sections used to distinguish the overlapping and gap areas are generated into a triangular mesh constraint surface between the two sections after being matched in order-preserving manner. All triangular facets are spliced ​​together to form the external spatial division domain Dex. This division domain serves as the contact / isolation interface between the ore bodies in the layers, which can accurately define the spatial boundary of the overlapping and gap areas of the ore bodies, and eliminate the defects of interlayer ore body interpenetration and boundary misalignment in traditional modeling.

[0168] In some embodiments, based on a dynamic programming algorithm with a prohibited point set, the shortest connection path that does not intersect with the internal spatial partitioning domain and the external spatial partitioning domain is extracted between the starting anchor point and the ending anchor point of the segment, including:

[0169] For each segment, the starting and ending anchor points of the segment are used as the endpoints that must be passed through, and the contour point sequence on the first profile and the contour point sequence on the second profile within the segment are used as the set of points to be matched.

[0170] Based on the point distance cost function and the forbidden point set, a dynamic programming recursive function is constructed. The point distance cost function is used to characterize the spatial distance between any contour point on the first profile and any contour point on the second profile. The forbidden point set is added to the forbidden point set if the triangular facet generated by connecting any contour point on the first profile and any contour point on the second profile intersects with any spatial segmentation domain.

[0171] Based on path functions and dynamic programming recursive functions, the shortest connection path that does not intersect with the internal or external spatial partitioning domains is extracted between the starting and ending anchor points of a segment. The path function is a sequence of point pairs from the starting anchor point to the ending anchor point, satisfying monotonicity, continuity, and mandatory passage through both the starting and ending anchor points. The dynamic programming recursive function calculates the cost by comparing the cumulative costs of three paths—diagonal, upward, and leftward—to the current position, and includes legal constraints: the legal constraints are that newly generated triangular faces during migration do not intersect with any spatial partitioning domain, and the intermediate points are not within the forbidden point set Ω; if a migration in a certain direction does not meet the legal constraints, the cost of that direction is set to +∞.

[0172] In some embodiments, the path function is:

[0173]

[0174] in, Let be the path function, representing the optimal connection scheme from the starting anchor point to the ending anchor point. For the first on the path The nth point pair represents the nth point of the first profile. The first contour point and the second profile of the second section Connect the contour points The path point number. This represents the total number of path pairs;

[0175] The dynamic programming recursive function is:

[0176] ;

[0177] in, To start from the anchor point Go to current location The cumulative minimum cost value, For the cost of distance between points, From the starting point to The cumulative cost, From the starting point to The cumulative cost, From the starting point to The cumulative cost, , To obtain the minimum value among the three paths, a valid path is defined as one where the newly generated triangular facets during the migration process do not intersect with any spatial partitioning domain, and the intermediate points are not in the forbidden point set. Inside; otherwise, the value in that direction is set to... .

[0178] The embodiments of this application will now be described in detail with reference to an application example.

[0179] The root of the problem with existing technologies lies in the lack of a unified mechanism to describe the spatial constraint relationships between ore bodies and to enable the surface models of each ore body to be generated collaboratively under this mechanism.

[0180] Based on this, this application example aims to provide a collaborative modeling method for multiple ore body contours, which can automatically identify overlaps, gaps, and elongated structures between ore body contours, construct spatial segmentation domains as constraints, and generate a three-dimensional model of multiple ore bodies that conforms to geological understanding under these constraints.

[0181] Below, in conjunction with Figure 7 The overall process of the technical solution in this application example is introduced below, with the following steps:

[0182] Step 701: Input data and parameter definition.

[0183] Let the two adjacent cross sections be respectively and .profile The area containing the outlines of two ore bodies is denoted as a closed polygon. , ;profile The corresponding outline is denoted as , Each contour line is a set of points arranged counterclockwise in the plane:

[0184] ;

[0185] Define the following threshold parameter:

[0186] -Overlap detection tolerance Distance threshold for determining whether two sides are collinear.

[0187] - Gap detection threshold : Determine the maximum distance between two contours to determine whether they constitute a gap.

[0188] - Narrow width threshold Areas smaller than this width are considered as narrow and elongated regions.

[0189] - Narrow shape factor threshold Used for preliminary determination of the overall narrow length.

[0190] Step 702: Data preprocessing: Extract planar dividing lines (overlapping dividing lines, gap skeleton lines, and narrow skeleton lines) from each group of ore body contours.

[0191] (1) Extraction of overlapping parts between contours

[0192] Within the same cross section and Iterate through all pairs of sub-segments of the two contours, i.e., line segments. With line segment Accurately determine the overlapping relationship of line segments based on distance constraints, and adopt a preset overlap determination tolerance. As a threshold for judgment. If a pair of line segments simultaneously meets both of the following conditions, then the two line segments are judged to have local overlap: First, the line segments... From the two vertices to the line segment or the perpendicular distance of its extension and line segment From the two vertices to the line segment The perpendicular distance between the line or its extension is less than the overlap determination tolerance. Second , There exists at least one line segment whose vertex has a perpendicular distance from another original line segment (excluding its extension) less than 1 / 2. For line segment pairs determined to be overlapping, the overlapping segments are further extracted, and all consecutive overlapping segments are merged to form an ordered point sequence, ultimately yielding the overlapping plane dividing line, denoted as... .

[0193] (2) Identification of narrow regions and extraction of skeleton lines

[0194] First, calculate the overall shape factor of the single ore body outline. ( The area enclosed by the outline. (where is the perimeter of the outline) The overall outline of the ore body is determined to be elongated. To accurately identify the elongated structure of local parts of the ore body, constrained triangulation is performed on the closed ore body outline. The triangulation satisfies the following topological rules: only the triangulation triangles located at both ends of the outline have two sides that fit the ore body outline, all other internal triangles have only one side that fits the ore body outline, and the remaining two sides of the triangles are all located within the ore body outline, ensuring that the triangulation structure fits the topological features of the ore body boundary. After completing the triangulation, the elongation attribute of each triangulation triangle is determined: for a single triangle, the side that fits the ore body outline is extracted as the reference side, and the vertical distance from the vertex of the triangle opposite the reference side to the reference side is calculated. This vertical distance is used as the basis for determining the local width of the triangle. If this vertical distance is less than a preset elongation width threshold... If the triangle is narrow, it is marked as such. Connect all adjacent narrow triangles to form a complete and continuous narrow domain of the ore body. Based on the topological relationship between the narrow domain and the ore body outline, the narrow domain is divided into three types: fully enclosed, semi-open, and fully open, and preprocessing is performed accordingly.

[0195] A fully enclosed, narrow domain: such as Figure 2 As shown, the fully enclosed narrow domain is a continuous, narrow, closed region corresponding to the entire ore body outline. Within this region, there are two characteristic triangles, which possess the topological characteristic that two sides conform to the ore body outline. The common vertex of the two outline sides of this characteristic triangle is located, and this vertex is used as the cutoff point to break the completely closed narrow domain. This decomposes the originally closed narrow domain into two independent polylines that are not connected to each other, completing the pre-processing of the fully enclosed narrow domain.

[0196] Semi-open, narrow domain: such as Figure 3As shown, the semi-open, narrow region is a local, narrow, non-closed area at the end of the ore body's outline. Within this region, there exists a unique characteristic triangle, which also possesses the topological characteristic that its two sides conform to the ore body's outline. By locking the common vertex of the two outline sides of this characteristic triangle, and using this as the truncation point, the semi-open, narrow region is disassembled and broken down, ultimately splitting the semi-open, narrow region into two independent, unconnected polyline structures.

[0197] Fully open narrow domain: such as Figure 4 As shown, the fully open narrow domain consists of two isolated and unconnected narrow line segments distributed in the middle area of ​​the ore body outline. These segments are themselves two independent and unconnected polyline structures, requiring no disassembly or preprocessing operations, and can be directly involved in the subsequent triangular network reconstruction and skeleton extraction process.

[0198] After unified preprocessing of the three types of narrow domains, all narrow domains are transformed into two independent, disconnected polyline structures. For the decomposed bi-segment polylines, an open segment wireframe algorithm is used to construct a triangulated network. Based on the reconstructed triangulated network, the edge midpoint concatenation method is used to extract the central skeleton of the narrow domain. The complete algorithm flow is as follows:

[0199] Step 1: Reconstruction of Triangulated Networks in Narrow Domains. For the two independent polylines formed after preprocessing, an open segment wireframe algorithm is used to construct a triangulated network.

[0200] Step 2: Extracting the midpoint of the skeleton feature. Traverse all the segmented triangles in the reconstructed triangular mesh, filter out all internal edges that do not conform to the ore body outline, and calculate the coordinates of the center midpoint of each internal edge.

[0201] Step 3: Ordered Serialization of Skeleton Lines. Based on the topological extension direction of the narrow domain, all extracted internal edge midpoints are arranged in an ordered and continuous sequence to form a smooth and continuous skeleton path that fits the geometric center of the narrow domain. The final continuous and ordered chain of nodes is the narrow planar dividing line, denoted as... The dividing line perfectly matches the center trajectory of the narrow domain and strictly adheres to the mandatory endpoint constraints, making it suitable for three types of narrow domain scenarios: fully enclosed, semi-open, and fully open.

[0202] (3) Gap region identification and segmentation line generation

[0203] For two sets of ore body outlines that do not overlap and are adjacent to each other within the same profile and Iterate through all boundary sub-segment pairs and filter out the closest points whose distance is less than the gap detection threshold. The adjacent line segments are identified, and overlapping areas are removed to obtain pure fracture boundary segments. The effective fracture consists of two boundary segments belonging to two different ore bodies and isolated from each other, with a topological structure consistent with the fully open narrow domain.

[0204] For the bi-segment independent polyline structure of the narrow domain of the gap, an open segment wireframe algorithm is used to construct a triangular mesh. The narrow domain skeleton extraction algorithm is reused: the midpoints of all internal non-boundary edges of the triangular mesh are extracted as skeleton feature nodes. These nodes are then sequentially connected along the gap's topological extension direction to generate a smooth skeleton path that fits the geometric center of the gap. The resulting ordered chain of nodes is the gap plane dividing line, denoted as […]. .

[0205] In summary, the overlapping planar segmentation lines, elongated planar segmentation lines, and gap planar segmentation lines extracted from this profile together constitute a complete set of profile segmentation lines. The overlapping planar segmentation lines and elongated planar segmentation lines are uniformly arranged in a counterclockwise spiral order according to the ore body outline, while the elongated planar segmentation lines are arranged in an orderly manner according to the topological extension direction of the elongated domain. This provides regular input data for subsequent cross-profile order-preserving structure matching and spatial segmentation domain construction.

[0206] Step 703: Spatial segmentation domain construction: Establish order-preserving structure matching between the planar segmentation lines of different profiles to connect and generate the internal and external spatial segmentation domains of the ore body.

[0207] (1) Order-preserving structure matching method

[0208] Given a sequence of planar dividing lines of two corresponding ore bodies and An optimal, order-preserving correspondence needs to be established between the two, allowing some elements to be skipped without changing their order.

[0209] Define the similarity matrix ,in

[0210] ;

[0211] In the formula For length similarity (using the smaller value of the ratio), The proximity is a Gaussian function based on the spatial distance between the midpoints. For directional consistency (dot product of unit direction vectors). These are the weighting coefficients.

[0212] Here, dynamic programming is used to solve for the maximum cumulative similarity:

[0213] ;

[0214] Backtracking yields the set of optimal matching pairs. The sequence is monotonically increasing. This method is called order-preserving structure matching.

[0215] (2) Construction of internal space segmentation domain (for narrow and long scenes)

[0216] like Figure 5 As shown, if and All are elongated ore bodies. Order-preserving structural matching is performed on their elongated planar segmentation sequence to obtain matching pairs. Each matching pair... and An open segment wireframe algorithm is used to construct triangular mesh surfaces. All combinations of triangular mesh surfaces constitute the internal spatial segmentation domain of the ore body. This is used to constrain the connection direction of the narrow and elongated parts of the ore body.

[0217] (3) Construction of external space segmentation domain (for overlapping / gap scenarios)

[0218] like Figure 6 As shown, if the cross-section and The surface contains planar dividing lines caused by overlap or gaps. These dividing line sequences are matched in a sequence-preserving manner, and the matched dividing lines are connected to generate a triangular mesh surface, which constitutes the external spatial division domain of the ore body. This segmentation domain directly defines the contact or isolation interface between the two ore bodies, and is a constraint surface that must be strictly controlled in subsequent modeling.

[0219] Step 704: Constrained Elastic Synchronous Mapping Modeling: Using the spatial segmentation domain as a hard constraint, adjacent contours are connected through elastic synchronous mapping to generate a multi-ore body 3D wireframe model.

[0220] This step uses the external / internal spatial segmentation domain as a hard constraint and generates a connecting triangular network between adjacent contours of the same ore body based on elastic synchronous mapping.

[0221] (1) Forced corresponding feature point extraction

[0222] set The endpoints of all elements are used as forced anchor points on the contour. For example, if the endpoints of overlapping dividing lines... belong Then it is in The corresponding point is given by the matching segmentation region. Arrange these anchor points in the order of the outline to divide the outline into multiple one-to-one corresponding segments.

[0223] (2) Constrained elastic synchronous mapping algorithm

[0224] For each corresponding segment, let the starting anchor point be... The final anchor point is The outline point series are respectively , Define the cost of distance between points. .

[0225] Define path It must satisfy the conditions of being monotonic, continuous, and passing through an anchor point. A forbidden point set is also introduced. If connected and If the resulting triangular patch intersects with any spatial partitioning domain, then Add to the prohibited set.

[0226] The dynamic programming recurrence formula is:

[0227] ;

[0228] "Legal" here means that the newly generated triangular facets during the migration process do not intersect with any spatial partitioning domain, and the midpoint is not located in the region specified in the original text. Inside; otherwise, the value in that direction is set to... Initial conditions The path is forced to pass through the anchor point.

[0229] (3) Constraint handling details

[0230] When both ends of a certain segment are overlapping partitioned domains, the surface of the ore body corresponding to that segment directly adopts the surface patch of that spatial partitioned domain, without the need for mapping generation, ensuring that the two ore body models fit perfectly at that point, without overlap or gaps.

[0231] Thus, based on the above technical solution, the following technical effects can be achieved: (1) High model fitting accuracy, completely eliminating overlapping gap defects: In the traditional modeling method, the triangular mesh connection of the contact position of multiple ore bodies is inconsistent, resulting in gaps in the model and mutual penetration and overlap of ore bodies. This application example constructs an external space segmentation domain to forcibly define the precise contact interface of multiple ore bodies, unifies the triangular mesh topology of the contact surface of adjacent ore bodies, eliminates the fitting deviation of the coplanar area of ​​ore bodies from the root of the algorithm, and realizes that the contact position of multiple ore bodies is completely fitted, without overlap and without gaps, perfectly solving the fitting distortion problem of traditional modeling. (2) Effectively avoids model interference and self-intersection, and has strong anti-distortion ability: In view of the technical drawbacks of traditional modeling, such as the lack of spatial constraints on adjacent ore bodies, the tendency of triangular meshes to cross boundaries and intrude into the space of adjacent ore bodies, and the easy distortion and self-intersection of the connection of narrow ore body contours, this application example constructs an internal spatial segmentation domain by decomposing and extracting the midpoint skeleton of the narrow domain, rigidly constraining the connection path of the narrow ore body contour, and isolating the space of adjacent ore bodies by the external segmentation domain. This not only avoids model interference between adjacent ore bodies, but also completely solves the distortion and self-intersection problems in the modeling of narrow ore bodies, and greatly improves the topological rationality of complex ore body models. (3) Full-process automated adaptation to complex geological scenarios: Traditional modeling often relies on manual intervention to correct overlapping, gaps, and narrow distortion problems, which is inefficient and highly subjective. This application example achieves full automation of the process, including automatic identification of overlapping ore body contours, gaps, and narrow structures; automatic extraction of planar dividing lines; cross-section order-preserving matching; and constraint mapping modeling. It eliminates the need for manual correction of various model defects and can adapt to a variety of complex geological conditions such as multiple ore body contacts, adjacent gaps, and narrow and complex ore bodies, significantly improving modeling efficiency and versatility.

[0232] Based on the hardware implementation of the above program modules, and in order to implement the method of the embodiments of this application, the embodiments of this application also provide an electronic device. Figure 8 The diagram shows only an exemplary structure of the electronic device, not the entire structure; implementation is possible as needed. Figure 8 The diagram shows part or all of the structure. For example... Figure 8 As shown, the electronic device 800 provided in this embodiment includes at least one processor 801, a memory 802, a user interface 803, and at least one network interface 804. The various components in the electronic device 800 are coupled together via a bus system 805. It can be understood that the bus system 805 is used to implement communication between these components. In addition to a data bus, the bus system 805 also includes a power bus, a control bus, and a status signal bus. However, for clarity, in... Figure 8 The general labeled all buses as Bus System 805.

[0233] The user interface 803 may include a monitor, keyboard, mouse, trackball, click wheel, buttons, touchpad, or touch screen.

[0234] The memory 802 in this embodiment is used to store various types of data to support the operation of the electronic device. Examples of such data include any computer program used to operate on the electronic device.

[0235] The multi-ore body contour modeling method for electronic devices disclosed in this application can be applied to or implemented by processor 801. Processor 801 may be an integrated circuit chip with signal processing capabilities. During implementation, each step of the multi-ore body contour modeling method for electronic devices can be completed by integrated logic circuits in the hardware of processor 801 or by instructions in software form. The processor 801 can be a general-purpose processor, a digital signal processor (DSP), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Processor 801 can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this application. A general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in the embodiments of this application can be directly manifested as execution by a hardware decoding processor, or execution by a combination of hardware and software modules in the decoding processor. The software modules can be located in a storage medium, specifically memory 802. Processor 801 reads information from memory 802 and, in conjunction with its hardware, completes the steps of the multi-ore body contour modeling method for electronic devices provided in the embodiments of this application.

[0236] In an exemplary embodiment, the electronic device may be implemented by one or more application-specific integrated circuits (ASICs), DSPs, programmable logic devices (PLDs), complex programmable logic devices (CPLDs), field-programmable gate arrays (FPGAs), general-purpose processors, controllers, microcontrollers (MCUs), microprocessors, or other electronic components to perform the aforementioned methods.

[0237] It is understood that memory 802 can be volatile memory or non-volatile memory, or both. Non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), ferromagnetic random access memory (FRAM), flash memory, magnetic surface memory, optical disc, or compact disc read-only memory (CD-ROM). Magnetic surface memory can be disk storage, or volatile memory can be random access memory (RAM) used as an external cache. By way of example, but not limitation, many forms of RAM are available, such as Static Random Access Memory (SRAM), Synchronous Static Random Access Memory (SSRAM), Dynamic Random Access Memory (DRAM), Synchronous Dynamic Random Access Memory (SDRAM), Double Data Rate Synchronous Dynamic Random Access Memory (DDRSDRAM), Enhanced Synchronous Dynamic Random Access Memory (ESDRAM), SyncLink Dynamic Random Access Memory (SLDRAM), and Direct Rambus Random Access Memory (DRRAM).The memories described in the embodiments of this application are intended to include, but are not limited to, these and any other suitable types of memories.

[0238] In an exemplary embodiment, this application also provides a computer storage medium, specifically a computer-readable storage medium storing a computer program thereon. This computer program can be executed by a processor to complete the steps of the method described in this application embodiment. The computer-readable storage medium can be a memory such as ROM, PROM, EPROM, EEPROM, Flash Memory, magnetic surface memory, optical disc, or CD-ROM.

[0239] It should be noted that terms such as "first" and "second" are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence.

[0240] Furthermore, the technical solutions described in the embodiments of this application can be arbitrarily combined without conflict. The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A method for modeling the contour lines of multiple ore bodies, characterized in that, The method includes: Obtain multiple ore body contour lines and preset threshold parameters on adjacent cross sections, wherein the multiple ore body contour lines are a set of points arranged in counterclockwise order; Based on the preset threshold parameters and the multiple ore body contour lines corresponding to each profile in the adjacent profiles, a set of planar dividing lines for each profile in the adjacent profiles is determined; wherein, the preset threshold parameters include: overlap determination tolerance, elongated shape factor threshold, elongated width threshold, and gap identification threshold; the set of planar dividing lines includes: overlapping planar dividing lines, elongated planar dividing lines, and gap planar dividing lines; the set of planar dividing lines includes a first set of planar dividing lines for a first profile and a second set of planar dividing lines for a second profile adjacent to the first profile; The first set of planar segmentation lines and the second set of planar segmentation lines are respectively used as the first planar segmentation sequence and the second planar segmentation sequence. A similarity matrix is ​​established between the first planar segmentation sequence and the second planar segmentation sequence. The maximum cumulative similarity function constructed by the similarity matrix is ​​solved based on dynamic programming, and the optimal matching pair set is generated based on the path backtracking corresponding to the maximum cumulative similarity. Based on the optimal matching pair set, an internal spatial segmentation domain is constructed between the narrow planar segmentation lines of two adjacent profiles and an external spatial segmentation domain is constructed between the overlapping planar segmentation lines or gap planar segmentation lines of two adjacent profiles. The endpoints of each element in the optimal matching pair set are used as forced anchor points on the ore body contours of the first and second profiles. Based on the forced anchor points, each ore body contour is divided into multiple one-to-one corresponding segments. For each of the multiple segments, with the internal spatial segmentation domain and the external spatial segmentation domain as hard constraints, a dynamic programming algorithm with a prohibited point set is used to extract the shortest connection path between the starting anchor point and the ending anchor point of the segment that does not intersect with the internal spatial segmentation domain and the external spatial segmentation domain. Based on each of the shortest connection paths, a connection triangular network of each ore body is generated. Based on each of the connection triangular networks, a three-dimensional wireframe model of multiple ore bodies is constructed.

2. The method according to claim 1, characterized in that, The step of determining the set of planar dividing lines for each profile in the adjacent profiles based on the preset threshold parameter and the multiple ore body contour lines corresponding to each profile in the adjacent profiles includes: Based on the overlap determination tolerance, the overlapping plane dividing line between multiple ore body outlines in each of the adjacent profiles is determined; Based on the elongated shape factor threshold and elongated width threshold, elongated planar dividing lines between multiple ore body outlines in each of the adjacent profiles are determined; Based on the gap identification threshold, the gap plane dividing line between multiple ore body contours in each of the adjacent profiles is determined; Based on the overlapping plane dividing line, the narrow plane dividing line, and the gap plane dividing line, the set of plane dividing lines is generated; The step of determining the overlapping plane dividing line between multiple ore body contours in each of the adjacent profiles based on the overlap determination tolerance includes: For each of the adjacent profiles, traverse all sub-line segment pairs of the multiple ore body contours within the profile; Extract overlapping sub-segment pairs from all sub-segments that satisfy a first set condition and a second set condition; the first set condition is that the first perpendicular distance between the two vertices of the first sub-segment and the second sub-segment or its extension in the sub-segment pair is less than the overlap determination tolerance; the second set condition is that at least one vertex of the first sub-segment and the second sub-segment has a perpendicular distance less than the overlap determination tolerance; extract overlapping segments from the overlapping sub-segment pairs and merge consecutive overlapping segments; generate an overlapping plane segmentation line set for the profile based on each consecutive overlapping segment; generate the overlapping plane segmentation line set based on the overlapping plane segmentation line set for each profile. The step of determining the plane dividing line of the gap between multiple ore body contours in each of the adjacent profiles based on the gap identification threshold includes: For each of the adjacent profiles, traverse all boundary sub-segment pairs of multiple ore body contours within the profile, and extract the nearest neighbor segments between all boundary sub-segment pairs whose nearest point distance is less than the gap identification threshold. Remove the areas where the adjacent line segments have been determined to overlap, and generate gap boundary segments; Based on the narrow domain skeleton extraction algorithm, a triangular network is constructed for the gap boundary segment; based on the gap topological extension direction, the midpoints of all internal non-boundary edges in the triangular network are sequentially connected to generate the gap plane segmentation line.

3. The method according to claim 2, characterized in that, The step of determining the elongated planar dividing line between multiple ore body contours in each of the adjacent profiles based on the elongated shape factor threshold and the elongated width threshold includes: For each of the multiple orebody contours in each of the adjacent cross-sections, the overall shape factor of the orebody contour is obtained. If the overall shape factor is less than the elongated shape factor threshold, constrained triangulation is performed on the interior of the orebody contour to generate multiple constrained triangulation triangles. These multiple constrained triangulation triangles satisfy positional constraints, namely: among the multiple constrained triangulation triangles, only the triangulation triangles located at the two ends of the contour have two sides that fit the orebody contour, while all other internal triangles have only one side that fits the orebody contour. For each constrained triangulation triangle, the vertex opposite the edge that fits the orebody contour is used as the reference edge. The third perpendicular distance to the reference edge is determined. If the third perpendicular distance is less than the narrow width threshold, the triangle is determined to be a narrow triangle. All adjacent narrow triangles are connected to form a narrow domain of the ore body. Based on the topological relationship between the narrow domain of the ore body and the ore body outline, the type of the narrow domain of the ore body is determined. The types of the narrow domain include: fully enclosed narrow domain, semi-open narrow domain, and fully open narrow domain. The fully enclosed narrow domain contains two characteristic triangles, and the two sides of the two characteristic triangles fit the ore body outline. The semi-open narrow domain contains a unique characteristic triangle, and the two sides of the characteristic triangle fit the ore body outline. The fully open narrow domain does not contain a characteristic triangle whose two sides fit the ore body outline. Based on the type of the narrow domain of the ore body, a target narrow domain is generated; a triangular mesh is constructed for the target narrow domain; the midpoints of all internal edges in the triangular mesh that do not fit with the outline of the ore body are extracted, and the midpoints are connected in an orderly manner according to the topological extension direction to construct the narrow plane dividing line of the outline of the ore body. Based on the narrow planar segmentation lines of each of the ore body contour lines, the narrow planar segmentation lines of the profile are generated; The elongated plane dividing lines are generated based on the elongated plane dividing lines of each of the aforementioned cross-sections.

4. The method according to claim 3, characterized in that, The generation of the target narrow region based on the type of the ore body narrow region includes: If the type of the ore body narrow domain is a fully enclosed narrow domain or a semi-open narrow domain, then the ore body narrow domain is preprocessed to generate the target narrow domain; If the type of the ore body elongated region is a fully open elongated region, then the fully open elongated region is the target elongated region; Wherein, if the type of the narrow ore body domain is a fully enclosed narrow domain or a semi-open narrow domain, then the narrow ore body domain is preprocessed to generate the target narrow domain, including: If the type of the ore body elongated domain is a fully enclosed elongated domain, then the common vertex of the two characteristic triangles with two sides that fit the outline of the ore body within the fully enclosed elongated domain is used as the cut-off point to decompose the closed elongated domain into two independent polylines that are not connected to each other; based on each of the independent polylines that are not connected to each other, the pre-processed elongated domain is generated. If the type of the ore body elongated domain is a semi-open elongated domain, then the common vertex of the characteristic triangle whose only two sides in the semi-open elongated domain region conform to the topological features of the ore body outline is used as the cutoff point, and the semi-open elongated domain is decomposed into two independent polylines that are not connected to each other. Based on the two independent polylines that are not connected to each other, the pre-processed elongated domain is generated.

5. The method according to claim 1, characterized in that, The similarity matrix is ​​as follows: ； in, The i-th dividing line in the first planar segmentation sequence With the j-th dividing line in the second plane segmentation sequence Similarity between them For length similarity, For location proximity, The dot product of unit direction vectors. These are the weighting coefficients, This is the sequence of first plane dividing lines. This is the sequence of second-plane dividing lines; The maximum cumulative similarity function is: ； in, From the starting position to the current position The maximum cumulative similarity value, The i-th dividing line in the first planar segmentation sequence With the j-th dividing line in the second plane segmentation sequence The similarity between them.

6. The method according to claim 1, characterized in that, The construction of an internal spatial segmentation domain between narrow planar dividing lines of adjacent two sections and an external spatial segmentation domain between overlapping planar dividing lines or gap planar dividing lines of adjacent two sections, based on the optimal matching pair set, includes: If the ore bodies corresponding to the two adjacent profiles are both narrow ore bodies, then the matching pairs in the optimal matching pair set that correspond to the narrow plane dividing lines are taken as internal matching pairs. For each internal matching pair, the open segment wireframe algorithm is used to construct the two narrow plane dividing lines corresponding to the matching pair into a triangular mesh surface. The internal spatial segmentation domain is constructed based on each triangular mesh surface. If there are overlapping plane dividing lines or gap plane dividing lines on both adjacent cross-sections, then the matching pairs in the optimal matching pair set corresponding to the overlapping plane dividing lines or gap plane dividing lines are taken as external matching pairs; for each external matching pair, the two dividing lines corresponding to the matching pair are connected to generate a triangular mesh surface; based on each triangular mesh surface, the external spatial segmentation domain is constructed.

7. The method according to claim 1, characterized in that, The dynamic programming algorithm based on a set of prohibited points extracts the shortest connection path between the starting and ending anchor points of the segment that does not intersect with the internal and external spatial partitioning domains, including: For each segment, the starting anchor point and the ending anchor point of the segment are used as the endpoints that must be passed through, and the contour point sequence on the first profile and the contour point sequence on the second profile within the segment are used as the set of points to be matched. A dynamic programming recursive function is constructed based on the point distance cost function and the forbidden point set. The point distance cost function is used to characterize the spatial distance between any contour point on the first profile and any contour point on the second profile. The forbidden point set is formed when a triangular facet generated by connecting any contour point on the first profile and any contour point on the second profile intersects with any of the spatial partitioning domains. Based on the path function and the dynamic programming recursive function, the shortest connection path that does not intersect with the internal and external spatial partitioning domains is extracted between the starting and ending anchor points of the segment. The path function is a sequence of point pairs from the starting anchor point to the ending anchor point, satisfying monotonicity, continuity, and forcibly passing through the starting and ending anchor points. The dynamic programming recursive function is recursively calculated by comparing the cumulative costs of three paths to the current position from the diagonal, upward, and leftward directions, and includes a legal constraint: the legal constraint is that the newly generated triangular facets during the migration process do not intersect with any of the spatial partitioning domains, and the intermediate points are not within the forbidden point set Ω; if the migration in a certain direction does not satisfy the legal constraint, the cost of that direction is set to +∞.

8. The method according to claim 7, characterized in that, The path function is: ； in, Let be the path function, representing the optimal connection scheme from the starting anchor point to the ending anchor point. For the first on the path The nth point pair represents the nth point of the first profile. The first contour point and the second profile of the second section Connect the contour points The path point number. This represents the total number of path pairs; The dynamic programming recursive function is: ； in, To start from the anchor point Go to current location The cumulative minimum cost value, For the cost of distance between points, From the starting point to The cumulative cost, From the starting point to The cumulative cost, From the starting point to The cumulative cost, , To obtain the minimum value among the three paths, a valid path is defined as one where the newly generated triangular facets during the migration process do not intersect with any spatial partitioning domain, and the intermediate points are not in the forbidden point set. Inside; otherwise, the value in that direction is set to... .

9. An electronic device, characterized in that, include: A processor and a memory for storing a computer program capable of running on the processor, wherein the processor, when running the computer program, performs the steps of the method as described in any one of claims 1 to 8.

10. A computer storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1 to 8.

Images