A method, device, system and storage medium for quickly modeling a three-dimensional ore body

Abstract

This application relates to the field of 3D modeling technology, specifically to a method, apparatus, system, and storage medium for rapid 3D modeling of ore bodies. The method includes: analyzing the distance and density differences between point cloud data from each station of the ore body, determining the spatial distribution anomaly coefficient, and processing noise in the point cloud data; performing surface fitting on the point cloud data of overlapping regions from each station, determining the non-smoothness feature index based on the similarity and curvature distribution characteristics between the point cloud data normal vectors, determining the overall structural complexity coefficient of each overlapping region by combining density distribution characteristics, correcting the number of sampling points in the registration algorithm, and matching the point cloud data of each overlapping region to obtain complete ore body point cloud data for establishing an ore body point cloud model. This application aims to improve the quality and reliability of the final ore body model.

Description

Technical Field

[0001] This application relates to the field of 3D modeling technology, specifically to a method, apparatus, system, and storage medium for rapid 3D modeling of ore bodies. Background Technology

[0002] With the continuous development of 3D digital technology in the field of geological engineering, 3D modeling optimization has gradually become a key technology for improving the accuracy of geological information analysis. The application of this technology in geological engineering has a significant positive impact on improving project quality and efficiency.

[0003] When performing 3D modeling of surface ore bodies, a large amount of initial point cloud data must first be acquired using 3D laser scanning technology. However, dust, water mist, or smoke generated during open-pit mining operations can scatter or absorb laser pulses, leading to signal attenuation and the generation of stray points, thus creating noise. Furthermore, due to the complex morphology of ore bodies, their surfaces are often rough, covered with loose debris, or exhibit complex textures. These factors can also make the laser echo signal irregular, thereby affecting the registration and processing of the point cloud data.

[0004] Existing conventional 3D ore body modeling methods do not adjust the number of sampling points of the algorithm when matching overlapping parts using point cloud registration algorithms, based on the complex mining environment and ore body surface morphology. This leads to uneven data coverage, missing important features, or loss of details, thus affecting the overall accuracy and completeness of the model. Ultimately, the constructed 3D model cannot truly reflect the actual situation of the ore body, limiting its application value in subsequent analysis and decision-making. Summary of the Invention

[0005] In view of the above, it is necessary to provide a method, device, system and storage medium for rapid 3D modeling of ore bodies to solve the above problems.

[0006] In a first aspect, this application provides a method for rapid three-dimensional modeling of ore bodies, the method comprising:

[0007] Acquire 3D point cloud data for each station in the ore body and define overlapping areas;

[0008] Analyze the distance differences and density differences between the point cloud data of each station and the overall distribution of its neighboring point cloud data to determine the spatial distribution anomaly coefficient of each point cloud data; and perform noise processing on the point cloud data based on the spatial distribution anomaly coefficient.

[0009] Analyze the denoised point cloud data, perform surface fitting on the point cloud data of each station belonging to the overlapping area, and determine the non-smooth feature index of each point cloud data in the overlapping area based on the similarity and curvature distribution characteristics between the normal vectors of the point cloud data and the neighboring point cloud data in the fitting surface.

[0010] The non-smooth feature index of each point cloud data in the overlapping region and its corresponding density distribution characteristics are analyzed to determine the overall structural complexity coefficient of each overlapping region. The number of sampling points in the registration algorithm is corrected, and the point cloud data of each overlapping region is matched using the registration algorithm to obtain complete ore body point cloud data and establish an ore body point cloud model.

[0011] Preferably, determining the spatial distribution anomaly coefficient for each point cloud data specifically involves:

[0012] The set of N nearest neighbor point cloud data for each point cloud data is taken as the local point cloud set of each point cloud data, where N is a preset value;

[0013] Obtain the fitted surface of the local point cloud set for each point cloud data. Based on the distance between each point cloud data and its corresponding fitted surface, and combined with the average distance between each point cloud data and all points in its local point cloud set, determine the degree of deviation of each point cloud data.

[0014] Based on the negative correlation mapping result of the average distance of each point cloud data, the local density of each point cloud data is obtained; the proportion of the difference in local density between each point cloud data and all point clouds in its local point cloud set is analyzed in the local density of each point cloud data, and the local density difference coefficient of each point cloud data is determined.

[0015] The deviation degree obtained from each point cloud data is positively fused with the local density difference coefficient to obtain the spatial distribution anomaly coefficient of each point cloud data.

[0016] Preferably, the noise processing of the point cloud data specifically includes:

[0017] For each station, the mean and standard deviation of the spatial distribution anomaly coefficient of all point cloud data are obtained, and the sum of the mean and the standard deviation of a preset multiple is used as the point cloud anomaly threshold; point cloud data with spatial distribution anomaly coefficients greater than the point cloud anomaly threshold are discarded as noisy point cloud data.

[0018] Preferably, determining the non-smooth feature index of each point cloud data in the overlapping region specifically involves:

[0019] Analyze the overall distribution characteristics of the normal vector similarity between each point cloud data in the overlapping region and all point cloud data in the local point cloud set in the overlapping region, and determine the orientation consistency coefficient of each point cloud data.

[0020] The negative correlation mapping result of the orientation consistency coefficient of each point cloud data in the overlapping region is positively fused with the curvature to obtain the non-smooth feature index of each point cloud data in the overlapping region.

[0021] Preferably, determining the overall structural complexity coefficient of each overlapping region specifically involves:

[0022] The product of the non-smoothness feature index and the local density difference coefficient of each point cloud data is used as the first feature value of the local structural complexity of each point cloud data.

[0023] By analyzing the dispersion of all the first eigenvalues ​​obtained in the overlapping region, the overall structural complexity coefficient of the overlapping region is obtained.

[0024] Preferably, the correction of the number of sampling points in the registration algorithm is specifically as follows:

[0025] The normalized result of the overall structural complexity coefficient of each overlapping region is multiplied by the upper limit of the range of the number of sampling points of the 4PCS algorithm to obtain the optimized value of the number of sampling points of the 4PCS algorithm in the corresponding overlapping region.

[0026] Preferably, the establishment of the ore body point cloud model includes:

[0027] The point cloud data is converted into a polygonal mesh using the Poisson reconstruction algorithm, and boundary holes are identified based on the topology of the mesh. Radial basis functions are used to repair the holes, resulting in a point cloud model of the ore body.

[0028] Secondly, this application provides a rapid 3D modeling system for ore bodies, applied to a rapid 3D modeling method for ore bodies, the system comprising:

[0029] The ore body point cloud acquisition module is used to acquire the three-dimensional point cloud data of each station of the ore body and define the overlapping area;

[0030] The ore body point cloud denoising module is used to analyze the distance difference and density difference characteristics between the point cloud data of each station and the overall distribution of its neighboring point cloud data, and to determine the spatial distribution anomaly coefficient of each point cloud data; based on the spatial distribution anomaly coefficient, noise processing is performed on the point cloud data.

[0031] The ore body point cloud analysis module is used to analyze the denoised point cloud data. It performs surface fitting on the point cloud data of each station in the overlapping area. Based on the similarity and curvature distribution characteristics between the normal vectors of the point cloud data and the neighboring point cloud data in the overlapping area in the fitted surface, it determines the non-smooth feature index of each point cloud data in the overlapping area.

[0032] The ore body point cloud model building module is used to analyze the non-smooth feature index of each point cloud data in the overlapping area and its corresponding density distribution characteristics, determine the overall structural complexity coefficient of each overlapping area, correct the number of sampling points in the registration algorithm, and use the registration algorithm to match the point cloud data of each overlapping area to obtain complete ore body point cloud data and build an ore body point cloud model.

[0033] Thirdly, this application provides a rapid three-dimensional modeling device for ore bodies, including a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor executes the computer program to implement the steps of any of the methods described above.

[0034] Fourthly, this application provides a storage medium for rapid three-dimensional modeling of ore bodies, on which computer program instructions are stored, which, when executed by a processor, implement the steps of the method described in any of the above-mentioned embodiments.

[0035] The beneficial effects of the above scheme are as follows: This application obtains three-dimensional point cloud data of each station of the ore body and defines overlapping areas. By obtaining the three-dimensional point cloud data of each station, the spatial structure of the ore body can be fully understood. At the same time, defining overlapping areas helps to ensure the continuity and consistency of data between different stations, laying the foundation for subsequent data processing and integration.

[0036] Secondly, the distance and density differences between the point cloud data of each station and the overall distribution of neighboring point cloud data are analyzed to determine the spatial distribution anomaly coefficient. This process can identify potential outliers and noise points. By calculating the spatial distribution anomaly coefficient, the representativeness and reliability of each point cloud data can be effectively evaluated, thus laying the foundation for subsequent noise processing and ensuring data quality. Noise processing based on the spatial distribution anomaly coefficient can significantly improve the accuracy of point cloud data, remove unnecessary interference information, and make subsequent modeling and analysis more accurate. This will help improve the quality and credibility of the final model.

[0037] Next, the denoised point cloud data is analyzed, and surface fitting is performed on the point cloud data of the overlapping area of ​​each station. This can better describe the geometric characteristics of the point cloud data and help to understand the shape and structure of the ore body. Based on the distribution characteristics of normal vector similarity and curvature, non-smooth areas in the point cloud data are effectively identified, and the non-smooth feature index of each point cloud data in the overlapping area is determined. This can reveal the structural complexity and irregularity in the point cloud data, which is of guiding significance for understanding the internal structure of the ore body and identifying special geological phenomena.

[0038] Finally, by combining the non-smooth feature index and density distribution characteristics, more comprehensive information about the ore body structure can be obtained, which helps to optimize the sampling strategy and registration algorithm, improve the accuracy and detail of the ore body model, correct the number of sampling points in the registration algorithm, and use the registration algorithm for matching. Adjusting the number of sampling points based on the previous analysis results can improve the efficiency and accuracy of registration, ensure the integrity and consistency of the ore body point cloud data, and construct a high-quality 3D model that can truly reflect the geological characteristics of the ore body, providing a solid foundation for subsequent analysis, decision-making and application. Attached Figure Description

[0039] Figure 1 A flowchart illustrating the steps of a rapid 3D modeling method for ore bodies provided in this application;

[0040] Figure 2 A block diagram of a rapid 3D modeling system for ore bodies provided in this application. Detailed Implementation

[0041] In the description of the embodiments in this application, the words "exemplary," "or," and "for example" are used to indicate examples, illustrations, or descriptions. Any embodiment or design scheme described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design schemes. Specifically, the use of the words "exemplary," "or," and "for example" is intended to present the relevant concepts in a specific manner.

[0042] 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 in this application's specification is for the purpose of describing particular embodiments only and is not intended to be limiting of the application.

[0043] It should also be noted that the terms "first" and "second" in this application and its accompanying drawings are used to distinguish similar objects, rather than to describe a specific order or sequence. The methods disclosed in the embodiments of this application or the methods shown in the flowcharts include one or more steps for implementing the method. Without departing from the scope of this application, the execution order of multiple steps can be interchanged, and some steps can also be deleted.

[0044] In all division operations covered in this application, to prevent computer program crashes or invalid values ​​from being generated due to a denominator being zero, a preset minimum threshold can be set for the denominator. When the absolute value of the denominator is less than the preset minimum threshold, it is adjusted to the minimum threshold with the same sign, thereby ensuring the robustness and feasibility of the algorithm under extreme conditions.

[0045] Please see Figure 1The diagram illustrates a flowchart of a rapid 3D modeling method for ore bodies according to an embodiment of this application. The method includes the following steps:

[0046] Step 1: Obtain the 3D point cloud data of each station in the ore body and define the overlapping area.

[0047] This application utilizes a 3D laser scanner to acquire point cloud data of the ore body. Before data acquisition, on-site exploration is necessary based on the ore body's extent and shape to rationally plan the location of monitoring stations, ensuring full coverage and minimizing obstruction. Then, the laser scanner is set up and leveled at the selected monitoring station, and the equipment is started to scan, thereby acquiring 3D point cloud data for each station. The Open3D library is used to employ a radius search method to obtain overlapping point cloud data between adjacent monitoring stations. The overlap ratio of point cloud data acquired from adjacent stations is 30%, which can be adjusted by the implementer according to actual conditions; this application does not impose any restrictions on this.

[0048] Step 2: Analyze the distance difference and density difference characteristics between the point cloud data of each station and the overall distribution of its neighboring point cloud data, and determine the spatial distribution anomaly coefficient of each point cloud data; based on the spatial distribution anomaly coefficient, perform noise processing on the point cloud data.

[0049] After acquiring high-density 3D point cloud data from each station, the raw data cannot be directly used to construct a 3D model. This is because the scanning environment is complex, and interference from the equipment itself and external factors can cause a large amount of noise in the acquired point cloud data. Furthermore, a single station can only acquire point cloud data from a local perspective of the ore body, and the data from each station are in an independent instrument coordinate system. Registration processing is also required to fuse the point cloud data from these stations with overlapping areas.

[0050] First, noise analysis is performed on the point cloud data. Under good laser reflection conditions, the acquired point cloud data exhibits strong continuity, densely and uniformly adhering to the ore body surface, clearly reflecting the ore body's geometry. However, during actual scanning, it is susceptible to environmental dust, water mist, and uneven reflectivity, resulting in discrete points randomly distributed within the scanning space, intertwined with surface points. For example, dust interference or refraction effects at structural edges often generate noise points that are far from the main structure; in damp areas, these uneven reflectivity locations typically cause the density of local noisy point cloud data to deviate significantly from the local density of surrounding point cloud data, resulting in a large density difference characteristic for outliers.

[0051] Based on the above characteristics, noise point clouds are analyzed and identified. This embodiment takes the i-th point cloud data from all point clouds at a single station as an example. This embodiment uses the K-nearest neighbor search algorithm to obtain the N nearest neighbor points, forming a local point cloud set for the i-th point cloud data. The value of N ranges from [20, 25], and in this embodiment it is set to 25. Implementers can choose the appropriate value based on their actual situation. Effective point cloud data is usually uniformly attached to the surface of the ore body and approximates a plane within a small local area. Therefore, this application uses principal component analysis to obtain the fitted surface corresponding to the local point cloud set. Then, based on the Euclidean distance between the i-th point cloud data and its corresponding fitted surface, the average distance between the i-th point cloud data and all point clouds within its local point cloud set is obtained. The ratio of the obtained Euclidean distance to the average distance is used as the deviation degree of the i-th point cloud data, denoted as . The result The larger the value, the greater the distance between the point cloud data and the actual ore body surface. Furthermore, the distribution of effective point cloud data is relatively uniform, resulting in relatively small local density differences between point cloud data points. In contrast, noisy point clouds, due to their random distribution, exhibit significant differences in local density compared to surrounding point clouds. Therefore, the negative correlation mapping result of the average distance corresponding to the i-th point cloud data is obtained as the local density of the i-th point cloud. In this embodiment, the negative correlation mapping result of the variable is calculated using the reciprocal of the variable. The absolute value of the difference in local density between the i-th point cloud data and other point cloud data within its local point cloud set is calculated. The ratio of the mean of all absolute differences to the local density of the i-th point cloud data itself is determined as the local density difference coefficient of the i-th point cloud data, denoted as [equation missing]. The result It reflects the density difference characteristics between point cloud data.

[0052] Furthermore, based on the deviation degree of the i-th point cloud data from its local range and the local density difference coefficient, the spatial distribution anomaly coefficient of the i-th point cloud data is obtained: the deviation degree obtained from each point cloud data and the local density difference coefficient are positively fused to obtain the spatial distribution anomaly coefficient of each point cloud data; in this embodiment, the positive fusion of multiple variables adopts the multiplication calculation method.

[0053] First, the degree of deviation, measured by the spatial distance between each point cloud data point and the fitted plane, directly reflects the extent to which a point geometrically deviates from its local surface. Second, the distribution of valid point clouds is usually relatively uniform, while the occurrence of noisy point clouds is random and disrupts this uniformity. Based on this, the density difference characteristics between each point cloud data point and its neighboring points were analyzed. By fusing the anomaly indicators from these two dimensions, if a point cloud data point simultaneously exhibits significant spatial discrepancies and density distribution anomalies, the likelihood of that point cloud being a noisy point cloud is higher.

[0054] Therefore, denoising of point cloud data can be performed based on the obtained spatial distribution anomaly coefficients. Specifically, the mean and standard deviation of the spatial distribution anomaly coefficients corresponding to all point cloud data at a given station are calculated and denoted as follows: , In this embodiment, the point cloud anomaly threshold is set to a value of [value missing]. Point cloud data with spatial distribution anomaly coefficients greater than or equal to the point cloud anomaly threshold are considered noisy point cloud data, and all noisy point cloud data are removed. Its advantage is that it has better adaptability in the construction of ore body 3D models compared with conventional denoising algorithms, and avoids introducing errors in point cloud data registration, which would lead to inaccurate point cloud data registration.

[0055] Step 3: Analyze the denoised point cloud data, perform surface fitting on the point cloud data of each station belonging to the overlapping area, and determine the non-smooth feature index of each point cloud data in the overlapping area based on the similarity between the normal vectors of the point cloud data and the neighboring point cloud data in the fitting surface and the distribution characteristics of the curvature.

[0056] After acquiring the denoised point cloud data from each monitoring station, considering that each station can only obtain point cloud data from a local perspective of the ore body, registration processing is required to obtain a complete 3D model of the ore body and remove redundant point cloud data. Specifically, this application adopts the 4PCS registration processing algorithm. In highly dynamic and uncertain mining environments, rapid and efficient 3D model construction is necessary; complex spatial structures, such as roadways, goafs, and benches, provide identifiable geometric features that are beneficial for registration, but overly complex or self-similar structures directly affect the efficiency and accuracy of registration. Therefore, the spatial structural characteristics of the ore body are analyzed.

[0057] Taking point cloud data from an overlapping region at a certain monitoring station as an example, this application uses the moving least squares method to obtain the fitted surface of this part of the point cloud data. The surface of the ore body may contain obvious edges, uneven textures, or large areas of smooth planes. The more complex the spatial structure of the region, the greater the curvature of the corresponding fitted surface, and the more diverse the normal vector directions of the point cloud data within a local area. Based on this, the curvature and normal vector of each point cloud data in the overlapping region are calculated in its corresponding fitted surface. The curvature reflects the curvature characteristics of the corresponding location, and the normal vector direction is uniformly set to point from the inside of the ore body to the outside.

[0058] Next, taking the j-th point cloud data as an example, the cosine similarity coefficient between the j-th point cloud data and the corresponding normal vector of each point cloud data in the local point cloud set of the overlapping region is calculated. The mean of all the cosine similarity coefficients is taken as the orientation consistency coefficient of the j-th point cloud data, denoted as . The result This reflects the degree of consistency between the direction of the corresponding normal vector of the point cloud data and other point cloud data in its neighborhood.

[0059] Furthermore, the negative correlation mapping result of the orientation consistency coefficient of each point cloud data in the overlapping region is positively fused with the curvature to obtain the non-smooth feature index of each point cloud data in the overlapping region. This index is used to measure the degree of curvature of the surface of the ore body structure where the point cloud data is located and the significance of the diversity characteristics of the normal vector direction. In this embodiment, the negative correlation mapping of the variable is specifically the reciprocal of the variable. It should be noted that, in order to prevent the denominator from being 0 and to ensure that the obtained non-smooth feature index is a number greater than zero, a value greater than 1 needs to be added to the denominator. In this embodiment, the value is 1.05, and the implementer can adjust it according to the actual situation.

[0060] By analyzing the curvature of the fitted surface containing each point cloud data and the directional diversity of the normal vector, the complexity of the ore body structure surface is quantified. When a point cloud data simultaneously exhibits high curvature and low directional consistency, the local non-smoothness it reflects becomes more pronounced.

[0061] Step 4: Analyze the non-smooth feature index of each point cloud data in the overlapping region and its corresponding density distribution characteristics, determine the overall structural complexity coefficient of each overlapping region, correct the number of sampling points in the registration algorithm, use the registration algorithm to match the point cloud data of each overlapping region, obtain complete ore body point cloud data, and establish an ore body point cloud model.

[0062] Point cloud data at structurally complex locations such as ore body edges and uneven textures exhibits a larger spatial distribution anomaly coefficient compared to point cloud data at smooth planar locations. This is because laser scanning of these locations, influenced by the stretching axis or length effect, causes density differences in point cloud data near sharp edges. Therefore, a comprehensive assessment of the complex features of the ore body structure is conducted by further considering the density distribution characteristics of the point cloud data. Each point cloud data point in the overlapping region possesses a corresponding local density difference coefficient. The product of the non-smoothness feature index of the j-th point cloud data point and its local density difference coefficient is taken as the first characteristic value of the local structural complexity of that point cloud data point, denoted as . This reflects the complexity of the structure of the point cloud data. Then, to obtain the overall spatial structure state of the overlapping region, this application calculates the standard deviation of all the first feature values ​​obtained in the overlapping region as the overall structural complexity coefficient of the overlapping region; this value reflects the overall complexity of the ore body structure in the overlapping region at a certain station location. In other embodiments, the overall structural complexity coefficient can also be obtained by calculating the coefficient of variation and Shannon entropy of all the first feature values.

[0063] Furthermore, the number of sampling points in 4PCS directly affects the efficiency and accuracy of the registration results. Specifically, a smaller number of sampling points results in higher registration efficiency; conversely, a larger number of sampling points results in lower registration efficiency but higher registration accuracy. Therefore, if the overall structural complexity is high, a larger number of sampling points is needed to ensure the registration accuracy of the point cloud data; conversely, a smaller number of sampling points can be used to improve registration efficiency when the overall structural complexity is low. The range of sampling point values ​​for the 4PCS algorithm is... In this embodiment , The overall structural complexity coefficients corresponding to all overlapping regions are statistically analyzed. The obtained overall structural complexity coefficients are normalized using the minimization method. The normalized result corresponding to a specific overlapping region is denoted as L. The formula for calculating the optimized number of sampling points is as follows: To avoid low registration accuracy due to an insufficient number of sampling points, a lower limit for the number of sampling points is set. Therefore, the 4PCS algorithm is used to register the point cloud data of each adjacent station to obtain complete ore body point cloud data.

[0064] Finally, since the obtained complete ore body point cloud data contains some redundancy, it needs to be further simplified. This application uses a point-by-point forward method to simplify the point cloud, effectively reducing the data volume while ensuring accuracy. Subsequently, the Poisson reconstruction algorithm is used to convert the point cloud into a polygonal mesh, and boundary holes are identified based on the mesh's topology. Finally, radial basis functions are used to repair these holes, thus obtaining a more complete ore body point cloud model.

[0065] Based on the same concept as the method embodiments of this application, a rapid 3D modeling system for ore bodies is proposed, comprising:

[0066] The ore body point cloud acquisition module is used to acquire the three-dimensional point cloud data of each station of the ore body and define the overlapping area;

[0067] The ore body point cloud denoising module is used to analyze the distance difference and density difference characteristics between the point cloud data of each station and the overall distribution of its neighboring point cloud data, and to determine the spatial distribution anomaly coefficient of each point cloud data; based on the spatial distribution anomaly coefficient, noise processing is performed on the point cloud data.

[0068] The ore body point cloud analysis module is used to analyze the denoised point cloud data. It performs surface fitting on the point cloud data of each station in the overlapping area. Based on the similarity and curvature distribution characteristics between the normal vectors of the point cloud data and the neighboring point cloud data in the overlapping area in the fitted surface, it determines the non-smooth feature index of each point cloud data in the overlapping area.

[0069] The ore body point cloud model building module is used to analyze the non-smooth feature index of each point cloud data in the overlapping area and its corresponding density distribution characteristics, determine the overall structural complexity coefficient of each overlapping area, correct the number of sampling points in the registration algorithm, and use the registration algorithm to match the point cloud data of each overlapping area to obtain complete ore body point cloud data and build an ore body point cloud model.

[0070] One such rapid 3D modeling system for ore bodies has the following block diagram: Figure 2 As shown.

[0071] Based on the same concept as the method embodiments of this application, a rapid three-dimensional modeling device for ore bodies is proposed, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any of the methods described above.

[0072] Based on the same concept as the method embodiments of this application, a storage medium for rapid three-dimensional modeling of ore bodies is proposed, which stores computer program instructions that, when executed by a processor, implement the steps of the method described above.

[0073] It should be noted that the flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of the systems, methods, and computer program products according to embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. In some alternative implementations, the functions marked in the blocks may occur in a different order than that shown in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. In the descriptions corresponding to the flowcharts and block diagrams in the accompanying drawings, the operations or steps corresponding to different blocks may also occur in a different order than disclosed in the description; sometimes there is no specific order between different operations or steps. For example, two consecutive operations or steps may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. Each block in a block diagram and / or flowchart, and combinations of blocks in a block diagram and / or flowchart, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0074] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A method for rapid three-dimensional modeling of ore bodies, characterized in that, The method includes the following steps: Acquire 3D point cloud data for each station in the ore body and define overlapping areas; Analyze the distance differences and density differences between the point cloud data of each station and the overall distribution of its neighboring point cloud data to determine the spatial distribution anomaly coefficient of each point cloud data; and perform noise processing on the point cloud data based on the spatial distribution anomaly coefficient. Analyze the denoised point cloud data, perform surface fitting on the point cloud data of each station belonging to the overlapping area, and determine the non-smooth feature index of each point cloud data in the overlapping area based on the similarity and curvature distribution characteristics between the normal vectors of the point cloud data and the neighboring point cloud data in the fitting surface. The non-smooth feature index of each point cloud data in the overlapping region and its corresponding density distribution characteristics are analyzed to determine the overall structural complexity coefficient of each overlapping region. The number of sampling points in the registration algorithm is corrected, and the point cloud data of each overlapping region is matched using the registration algorithm to obtain complete ore body point cloud data and establish an ore body point cloud model. Specifically, determining the spatial distribution anomaly coefficient for each point cloud data involves: taking the set of N nearest neighbor point cloud data for each point cloud data as the local point cloud set for each point cloud data, where N is a preset value; obtaining the fitted surface of the local point cloud set for each point cloud data; determining the deviation degree of each point cloud data based on the distance between each point cloud data and its corresponding fitted surface, combined with the average distance between each point cloud data and all point clouds within its local point cloud set; obtaining the local density of each point cloud data based on the negative correlation mapping result of the average distance of each point cloud data; analyzing the proportion of the difference in local density between each point cloud data and all point clouds within its local point cloud set in the local density of each point cloud data to determine the local density difference coefficient for each point cloud data; and positively fusing the deviation degree obtained for each point cloud data with the local density difference coefficient to obtain the spatial distribution anomaly coefficient for each point cloud data. The determination of the overall structural complexity coefficient of each overlapping region is specifically as follows: the product of the non-smooth feature index and the local density difference coefficient of each point cloud data is used as the first feature value of the local structural complexity of each point cloud data; the dispersion of all the first feature values ​​obtained in the overlapping region is analyzed to obtain the overall structural complexity coefficient of the overlapping region.

2. The rapid 3D modeling method for ore bodies as described in claim 1, characterized in that, The noise processing of the point cloud data specifically includes: For each station, the mean and standard deviation of the spatial distribution anomaly coefficient of all point cloud data are obtained, and the sum of the mean and the standard deviation of a preset multiple is used as the point cloud anomaly threshold. Point cloud data with spatial distribution anomaly coefficients greater than the point cloud anomaly threshold are discarded as noisy point cloud data.

3. The rapid 3D modeling method for ore bodies as described in claim 1, characterized in that, The determination of the non-smooth feature index of each point cloud data in the overlapping region is specifically as follows: Analyze the overall distribution characteristics of the normal vector similarity between each point cloud data in the overlapping region and all point cloud data in the local point cloud set in the overlapping region, and determine the orientation consistency coefficient of each point cloud data. The negative correlation mapping result of the orientation consistency coefficient of each point cloud data in the overlapping region is positively fused with the curvature to obtain the non-smooth feature index of each point cloud data in the overlapping region.

4. The rapid three-dimensional modeling method for ore bodies as described in claim 1, characterized in that, The correction to the number of sampling points in the registration algorithm is specifically as follows: The normalized result of the overall structural complexity coefficient of each overlapping region is multiplied by the upper limit of the range of the number of sampling points of the 4PCS algorithm to obtain the optimized value of the number of sampling points of the 4PCS algorithm in the corresponding overlapping region.

5. The rapid three-dimensional modeling method for ore bodies as described in claim 1, characterized in that, The establishment of the ore body point cloud model includes: The point cloud data is converted into a polygonal mesh using the Poisson reconstruction algorithm, and boundary holes are identified based on the topology of the mesh. Radial basis functions are used to repair the holes, resulting in a point cloud model of the ore body.

6. A rapid 3D modeling system for ore bodies, applied to the rapid 3D modeling method for ore bodies according to claim 1, characterized in that, The system includes: The ore body point cloud acquisition module is used to acquire the three-dimensional point cloud data of each station of the ore body and define the overlapping area; The ore body point cloud denoising module is used to analyze the distance difference and density difference characteristics between the point cloud data of each station and the overall distribution of its neighboring point cloud data, and to determine the spatial distribution anomaly coefficient of each point cloud data; based on the spatial distribution anomaly coefficient, noise processing is performed on the point cloud data. The ore body point cloud analysis module is used to analyze the denoised point cloud data. It performs surface fitting on the point cloud data of each station in the overlapping area. Based on the similarity and curvature distribution characteristics between the normal vectors of the point cloud data and the neighboring point cloud data in the overlapping area in the fitted surface, it determines the non-smooth feature index of each point cloud data in the overlapping area. The ore body point cloud model building module is used to analyze the non-smooth feature index of each point cloud data in the overlapping area and its corresponding density distribution characteristics, determine the overall structural complexity coefficient of each overlapping area, correct the number of sampling points in the registration algorithm, and use the registration algorithm to match the point cloud data of each overlapping area to obtain complete ore body point cloud data and build an ore body point cloud model.

7. A rapid 3D modeling device for ore bodies, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method as described in any one of claims 1-5.

8. A storage medium for rapid three-dimensional modeling of ore bodies, characterized in that, It stores computer program instructions that, when executed by a processor, implement the steps of the method described in any one of claims 1-5.

Images