Data processing method and device, computer device and storage medium

Abstract

The application relates to a data processing method and device, computer equipment, a storage medium and a computer program product. The method comprises the following steps: obtaining three-dimensional structure data and resource exchange information of a historical object; performing feature extraction on the three-dimensional structure data to obtain feature data of the historical object; determining a feature vector of the historical object based on the feature data; storing the feature vector and the identifier of the historical object in a target database; obtaining a feature vector of a query object, matching a target historical object matching the feature vector of the query object from the target database according to the similarity relationship between the feature vector of the query object and the feature vector of each historical object; and generating reference resource exchange information of the query object according to the indexed resource exchange information of the historical object. The method can improve the accuracy and efficiency of resource exchange information generation.

Description

Technical Field

[0001] This application relates to the field of big data technology, and in particular to a data processing method, apparatus, computer equipment, and storage medium. Background Technology

[0002] With the rapid development of 3D digital design and manufacturing technologies, enterprises have accumulated massive amounts of 3D structural data during the production process. This data carries complete design information for components and is a crucial technological asset for enterprises. Simultaneously, historical production data related to these components is also stored in various business systems in multiple forms. In traditional technologies, when faced with new component requirements, enterprises typically rely on engineers manually reviewing historical drawings, subjectively comparing the similarity of component structures, and estimating the required resource consumption based on personal experience. This approach is highly dependent on human experience, resulting in low efficiency and inconsistent, unreliable estimations. Summary of the Invention

[0003] Therefore, it is necessary to provide a data processing method, apparatus, computer equipment, computer-readable storage medium, and computer program product to address the aforementioned technical problems.

[0004] Firstly, this application provides a data processing method. The method includes:

[0005] Acquire the 3D structural data and resource exchange information of historical objects;

[0006] Feature extraction is performed on the three-dimensional structural data to obtain the feature data of the historical object; wherein, the feature data includes at least one of the following: geometric features, topological features, and shape features;

[0007] Based on the feature data, the feature vector of the historical object is determined;

[0008] The feature vector is associated with the identifier of the historical object and stored in the target database;

[0009] Obtain the feature vector of the query object, and based on the similarity relationship between the feature vector of the query object and the feature vectors of each historical object, match the target historical object that matches the feature vector of the query object from the target database;

[0010] Based on the resource exchange information of the target historical object, reference resource exchange information of the queried object is generated; the resource exchange information is used to characterize the quantity of exchangeable resources of the object.

[0011] In one embodiment, the geometric feature includes at least one of the following: volume feature, surface area feature, and compact feature;

[0012] The acquisition of volume features includes: obtaining volume data of the historical object from the three-dimensional structural data to obtain volume features;

[0013] The acquisition of the surface area features includes: obtaining the surface area data of the historical object from the three-dimensional structural data to obtain the surface area features;

[0014] The acquisition of the compact features includes: acquiring the volume data and surface area data, and generating compact features that characterize the spatial composition of the historical objects based on the volume data and surface area data.

[0015] In one embodiment, the topological features include at least one of the following: face features, edge features, vertex features, and topological structure features;

[0016] The acquisition of the surface features includes: traversing the three-dimensional structural data, determining the number of surfaces contained in the historical object, and obtaining the surface features;

[0017] The acquisition of the edge features includes: traversing the three-dimensional structure data, determining the number of edges contained in the historical object, and obtaining the edge features;

[0018] The acquisition of point features includes: traversing the three-dimensional structural data, determining the number of points contained in the historical object, and obtaining point features;

[0019] The acquisition of the topological features includes: acquiring the number of faces, edges, and vertices of the historical object, and generating the topological features of the historical object based on the number of faces, edges, and vertices.

[0020] In one embodiment, the shape feature includes at least one of the following: a spherical feature, a filled feature, and an elongated feature;

[0021] The acquisition of the spherical feature includes: acquiring the volume of the historical object and the volume of the corresponding minimum circumscribed sphere, and generating the spherical feature of the historical object based on the volume of the historical object and the volume of the minimum circumscribed sphere;

[0022] The acquisition of the filling feature includes: acquiring the volume of the historical object and the volume of the corresponding minimum bounding cube, and generating the filling feature of the historical object based on the volume of the historical object and the volume of the minimum bounding cube;

[0023] The acquisition of the elongation feature includes: acquiring the length of the historical object in three principal axis directions, and generating the elongation feature of the historical object based on the correspondence between the lengths in the three principal axis directions.

[0024] In one embodiment, matching target historical objects from the target database that match the feature vector of the query object based on the similarity relationship between the feature vector of the query object and the feature vectors of each of the historical objects includes:

[0025] The feature vector of the query object is compared with multiple historical feature vectors stored in the target database to determine the similarity relationship between the feature vector of the query object and each of the historical feature vectors.

[0026] Based on the similarity relationship, target historical objects that match the feature vector of the query object are determined from the target database.

[0027] In one embodiment, the step of comparing the feature vector of the query object with multiple historical feature vectors stored in the target database further includes:

[0028] Retrieve the volume and length data of the queried object;

[0029] Based on the volume and length data of the queried object, historical feature vectors with volume or size differences less than a preset threshold are filtered from the target database to obtain candidate feature vectors.

[0030] In one embodiment, the feature data further includes at least one of the following: rotational features, outline features, symmetry features, and undulation features;

[0031] The acquisition of the rotation features includes: acquiring the rotation parameters of the historical object, and generating the rotation features of the historical object based on the rotation parameters;

[0032] The acquisition of the outline features includes: acquiring the volume of the minimum convex hull of the historical object and the volume of the historical object, and generating the outline features of the historical object based on the volume of the minimum convex hull and the volume of the historical object itself;

[0033] The acquisition of the symmetry features includes: detecting the symmetry relationship of the historical object in three orthogonal directions, and generating the symmetry features of the historical object based on the symmetry relationship;

[0034] The acquisition of the undulation features includes: acquiring the curvature data of the surface of the historical object and the distribution of the curvature data, and generating the undulation features of the historical object based on the curvature data and the distribution of the curvature data.

[0035] Secondly, this application also provides a data processing apparatus. The apparatus includes:

[0036] The data acquisition module is used to acquire the three-dimensional structural data and resource exchange information of historical objects;

[0037] The feature extraction module is used to extract features from the three-dimensional structural data to obtain feature data of the historical object; wherein, the feature data includes at least one of the following: geometric features, topological features, and shape features;

[0038] The feature extraction module is also used to determine the feature vector of the historical object based on the feature data;

[0039] The data storage module is used to associate the feature vector with the identifier of the historical object and store it in the target database;

[0040] The data indexing module is used to obtain the feature vector of the query object, and match the target historical object that matches the feature vector of the query object from the target database based on the similarity relationship between the feature vector of the query object and the feature vector of each historical object.

[0041] The data generation module is used to generate reference resource exchange information for the queried object based on the resource exchange information of the target historical object; the resource exchange information is used to characterize the quantity of exchangeable resources of the object.

[0042] In one embodiment, the geometric features include at least one of the following: volume features, surface area features, and compactness features; the feature extraction module is further configured to:

[0043] Volume data of the historical object is obtained from the three-dimensional structural data to obtain volume features;

[0044] The surface area data of the historical object is obtained from the three-dimensional structural data to obtain surface area features;

[0045] The volume data and surface area data are acquired, and a compact feature characterizing the spatial composition of the historical object is generated based on the volume data and surface area data.

[0046] In one embodiment, the topological features include at least one of the following: face features, edge features, vertex features, and topological structure features; the feature extraction module is further configured to:

[0047] Traverse the three-dimensional structural data to determine the number of faces contained in the historical object and obtain the face features;

[0048] Traverse the three-dimensional structure data to determine the number of edges contained in the historical object and obtain the edge features;

[0049] Traverse the three-dimensional structural data to determine the number of points contained in the historical object and obtain point features;

[0050] Obtain the number of faces, edges, and vertices of the historical object, and generate the topological features of the historical object based on the number of faces, edges, and vertices.

[0051] In one embodiment, the shape feature includes at least one of the following: a spherical feature, a filled feature, and an elongated feature; the feature extraction module is further configured to:

[0052] Obtain the volume of the historical object and the volume of the corresponding minimum circumscribed sphere, and generate the spherical feature of the historical object based on the volume of the historical object and the volume of the minimum circumscribed sphere;

[0053] Obtain the volume of the historical object and the volume of the corresponding minimum bounding box, and generate the filling feature of the historical object based on the volume of the historical object and the volume of the minimum bounding box;

[0054] The lengths of the historical object in the three principal axis directions are obtained, and the elongation features of the historical object are generated according to the correspondence between the lengths in the three principal axis directions.

[0055] In one embodiment, the data indexing module is further configured to:

[0056] The feature vector of the query object is compared with multiple historical feature vectors stored in the target database to determine the similarity relationship between the feature vector of the query object and each of the historical feature vectors.

[0057] Based on the similarity relationship, target historical objects that match the feature vector of the query object are determined from the target database.

[0058] In one embodiment, the apparatus further includes a data filtering module for:

[0059] Retrieve the volume and length data of the queried object;

[0060] Based on the volume and length data of the queried object, historical feature vectors with volume or size differences less than a preset threshold are filtered from the target database to obtain candidate feature vectors.

[0061] In one embodiment, the feature data further includes at least one of the following: rotational features, contour features, symmetry features, and undulation features; the feature extraction module is further configured to:

[0062] Obtain the rotation parameters of the historical object, and generate the rotation features of the historical object based on the rotation parameters;

[0063] Obtain the volume of the minimum convex hull of the historical object and the volume of the historical object itself, and generate the outline feature of the historical object based on the volume of the minimum convex hull and the volume of the historical object itself;

[0064] The symmetry relationship of the historical object in three orthogonal directions is detected, and the symmetry feature of the historical object is generated based on the symmetry relationship;

[0065] Obtain the curvature data of the surface of the historical object and the distribution of the curvature data, and generate the undulation features of the historical object based on the curvature data and the distribution of the curvature data.

[0066] Thirdly, this application also provides a computer device. The computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the data processing method as described in any one of the embodiments of this disclosure.

[0067] Fourthly, this application also provides a computer-readable storage medium. The computer-readable storage medium stores a computer program thereon, which, when executed by a processor, implements the data processing method as described in any one of the embodiments of this disclosure.

[0068] Fifthly, this application also provides a computer program product. The computer program product includes a computer program that, when executed by a processor, implements the data processing method as described in any of the embodiments of this disclosure.

[0069] The aforementioned data processing methods, apparatus, computer equipment, storage media, and computer program products extract multi-dimensional features from the three-dimensional structural data of historical objects, transforming difficult-to-quantify geometric forms into feature vectors containing at least one of geometric, topological, and shape features. They also establish an association storage mechanism between feature vectors and resource exchange information, constructing a target database with structural similarity as the core index. When faced with a new query object, this method eliminates the need for cumbersome cost accounting processes. Instead, it quickly retrieves historical objects with similar structural features based on similarity matching between feature vectors, and uses their actual resource exchange information as a reference to generate reliable reference resource exchange information. This data processing method effectively solves the technical problems of low data retrieval efficiency, delayed reference information acquisition, and large cost estimation deviations in traditional methods, significantly improving the accuracy and timeliness of resource exchange information evaluation. Attached Figure Description

[0070] Figure 1 This is a diagram illustrating the application environment of a data processing method in one embodiment.

[0071] Figure 2This is a flowchart illustrating a data processing method in one embodiment;

[0072] Figure 3 This is a flowchart illustrating the implementation of a data processing method in one embodiment;

[0073] Figure 4 This is a structural block diagram of a data processing device in one embodiment;

[0074] Figure 5 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0075] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0076] The data processing method provided in this application embodiment can be applied to, for example... Figure 1 In the application environment shown, terminal 102 communicates with server 104 via a network. A data storage system can store the data that server 104 needs to process. The data storage system can be integrated onto server 104 or placed on a cloud or other network server. Specifically, server 104 is responsible for core data processing and model calculations, obtaining 3D structural data and resource exchange information of historical objects from the data storage system, and extracting features from this 3D structural data to obtain feature data such as geometric features, topological features, and shape features of the historical objects. Based on this feature data, server 104 determines the feature vector of the historical object and associates this feature vector with the identifier of the historical object, storing it in the target database. When terminal 102 needs to evaluate the resource exchange value of a query object, the user collects or uploads the 3D data of the query object through terminal 102, which then sends it to server 104. After receiving the query object data, server 104 extracts its feature vector and performs index matching based on a pre-built target database to find the most similar historical object. Based on the resource exchange information of the indexed historical objects, server 104 generates reference resource exchange information for the query object and returns this information to terminal 102 for display. Terminal 102 can be, but is not limited to, various personal computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. IoT devices can include smart speakers, smart TVs, smart air conditioners, and smart in-vehicle devices. Portable wearable devices can include smartwatches, smart bracelets, and head-mounted devices. Server 104 can be implemented using a standalone server or a server cluster consisting of multiple servers.

[0077] In one embodiment, such as Figure 2 As shown, a data processing method is provided, including the following steps:

[0078] Step S200: Obtain the three-dimensional structural data and resource exchange information of historical objects.

[0079] The historical objects may include completed components, work-in-process in the structural stage, or prototypes in the design stage. The three-dimensional structural data of the historical objects can be obtained through three-dimensional model files generated during the design stage, such as STP, STEP, and STL files. The shape, size, and structure of the historical objects can be obtained from these three-dimensional model files. The resource exchange information may include various resource data actually consumed during the production process of the historical objects, such as raw material costs, processing time, energy consumption, and process costs. The resource exchange information may also include transaction data generated during the circulation of the historical objects, such as market quotations and transaction prices.

[0080] Step S202: Extract features from the three-dimensional structural data to obtain feature data of the historical object; wherein the feature data includes at least one of the following: geometric features, topological features, and shape features.

[0081] The feature data can be used to quantify the geometric and physical attributes of historical objects from multiple dimensions, specifically including one or more of the following types: geometric features, size features, topological features, shape features, rotational features, outline features, symmetry features, undulation features, etc.; geometric features can be used to describe the basic dimensions and space occupancy of the object, specifically including one or more of volume, surface area, compactness, etc.; size features can be used to describe the scale information of the object in three dimensions; specifically including one or more of length, width, height, etc.; topological features can be used to describe the surface and structural composition of the object; specifically including one of face number, edge number, vertex number, etc. One or more; shape features can be used to describe the overall shape and proportion of an object; specifically, they can include one or more of sphericity, rectangularity, elongation, flatness, etc.; rotational features can be used to describe the inertial properties of an object in rotational motion; specifically, they can include the moments of inertia of the three principal axes, etc.; contour features can be used to describe the fullness and concavity / convexity of an object's contour; specifically, they can include one or more of convexity, solidity, etc.; symmetry features can be used to describe the symmetrical distribution of an object in space; specifically, they can include mirror symmetry in three directions, etc.; undulation features can be used to describe the local changes and curvature distribution of an object's surface; specifically, they can include one or more of the mean curvature and standard deviation of curvature, etc.

[0082] In an exemplary embodiment, feature data extraction can be achieved through a predefined feature extraction algorithm or through automatic learning based on a deep learning model. Specifically, when using a deep learning model, a feature extractor based on a 3D convolutional neural network or point cloud network can be constructed. The 3D structural data of historical objects is converted into a unified representation, such as a voxel mesh or raw point cloud, as input to the model; and through multi-layer convolution or pooling operations, abstract feature representations of objects at different levels, such as local geometric structure, global shape contour, and detailed undulation features, are automatically learned.

[0083] Step S204: Based on the feature data, determine the feature vector of the historical object.

[0084] The feature vector can include standardizing and unifying the dimensions of various extracted feature data to form a numerical vector representation of a preset length. Specifically, different types of feature data, such as geometric features, topological features, and shape features, can be normalized to eliminate the influence of dimensional differences on similarity calculation; the processed feature data can then be concatenated according to a preset arrangement order to construct a comprehensive feature vector of the historical object.

[0085] In one exemplary embodiment, determining the feature vector may include combining various features to construct a high-dimensional vector. Specifically, it may include weighting and combining the values ​​of volume, surface area, and compactness in geometric features, the number of faces, edges, and vertices in topological features, and the sphericity, fill factor, and elongation in shape features according to preset weights to form a feature vector of fixed dimensions; or using dimensionality reduction methods such as principal component analysis to extract key principal components from the original high-dimensional features and construct a low-dimensional feature representation.

[0086] Step S206: Associate the feature vector with the identifier of the historical object and store it in the target database.

[0087] The target database can be a vector database or a database system that supports efficient similarity retrieval. The identifier of the historical object can include metadata such as object number, name, version information, and creation time to uniquely identify the historical object. Specifically, during storage, feature vectors can be used as indexes, and the identifier, feature vectors, three-dimensional structural data, and resource exchange information of the historical object can be stored as associated values ​​so that subsequent similarity retrieval based on feature vectors can quickly locate the corresponding historical object and its complete information.

[0088] In one embodiment, the construction of the target database may further include clustering analysis of historical feature vectors, grouping similar historical objects into the same category, and establishing a hierarchical index structure to improve retrieval efficiency in large-scale data scenarios. Specifically, hierarchical clustering or clustering algorithms based on density peaks can be used to divide the feature space into multiple sub-regions, each corresponding to a cluster center. During retrieval, the distance between the query object and each cluster center is calculated to quickly locate the most relevant sub-region, and a refined similarity comparison is performed within the corresponding sub-region, thereby significantly reducing computational load and accelerating indexing efficiency.

[0089] Step S208: Obtain the feature vector of the query object, and match the target historical object that matches the feature vector of the query object from the target database based on the similarity relationship between the feature vector of the query object and the feature vector of each historical object.

[0090] The query objects can include new components to be evaluated, prototype products in the design stage, or 3D models uploaded by users. The method of obtaining the feature vectors of the query objects can be the same as that of obtaining the feature vectors of historical objects. That is, the same feature extraction and vectorization processing is performed on the 3D structural data of the query objects to ensure the consistency of the feature space.

[0091] Specifically, the indexing process may include calculating the similarity measure between the feature vector of the query object and each historical feature vector in the target database, such as Euclidean distance, cosine similarity, or Mahalanobis distance; sorting the similarities from high to low, and selecting one or more historical objects with the highest similarity as the matching results. In an exemplary embodiment, an approximate nearest neighbor search algorithm, such as a method based on locality-sensitive hashing or product quantization, may also be used to significantly improve the retrieval speed of large-scale vector data while ensuring retrieval accuracy.

[0092] Step S210: Generate reference resource exchange information for the queried object based on the resource exchange information of the target historical object; the resource exchange information is used to characterize the quantity of exchangeable resources of the object.

[0093] The generation of reference resource exchange information can include various strategies such as similarity-based weighted fusion, nearest neighbor estimation, or regression model prediction. Specifically, when multiple historical objects are indexed, their resource exchange information can be weighted and averaged based on the feature similarity between each historical object and the query object, assigning greater weight to historical objects with higher similarity, thereby obtaining a more accurate reference estimate. Alternatively, the resource exchange information of the single historical object with the highest similarity can be directly selected as the reference value for the query object, which is suitable for scenarios where high accuracy is not required or where the data distribution is relatively concentrated. Furthermore, a mapping model based on feature vectors to resource exchange information can be constructed, such as support vector regression, random forests, or neural networks, using historical data to train model parameters to achieve direct prediction of the query object's resource exchange information.

[0094] In one exemplary embodiment, the generation of reference resource exchange information may further include uncertainty quantification and confidence interval estimation. Specifically, the confidence interval of the reference estimate can be calculated based on the number of matched historical objects, the similarity distribution, and the dispersion of the resource exchange information, providing users with decision-making references. When the resource exchange information of historical objects fluctuates significantly or the matching similarity is generally low, the credibility level of the reference resource exchange information can be marked to prompt users to use it with caution or collect more data.

[0095] In the aforementioned data processing method, multi-dimensional feature extraction is performed on the three-dimensional structural data of historical objects, transforming the difficult-to-quantify geometric forms into feature vectors containing at least one of geometric, topological, and shape features. A correlation storage mechanism between feature vectors and resource exchange information is established, constructing a target database with structural similarity as the core index. When faced with a new query object, this method no longer requires a cumbersome cost accounting process. Instead, based on the similarity matching between feature vectors, it quickly retrieves historical objects with similar structural features and uses their actual resource exchange information as a reference to generate reliable reference resource exchange information. This data processing method effectively solves the technical problems of low data retrieval efficiency, delayed reference information acquisition, and large cost estimation deviations in traditional methods, significantly improving the accuracy and timeliness of resource exchange information evaluation.

[0096] In one embodiment, the geometric feature includes at least one of the following: volume feature, surface area feature, and compact feature;

[0097] The acquisition of volume features includes: obtaining volume data of the historical object from the three-dimensional structural data to obtain volume features.

[0098] The volumetric feature can include the three-dimensional space occupied by the historical object in the design space, which is a key indicator for measuring the object's material consumption and basic cost. Specifically, the volume data can be obtained directly by parsing from the model file corresponding to the three-dimensional structural data, or by discretizing the three-dimensional model into a voxel mesh using a voxelization method, counting the number of voxels, and multiplying by the volume of a single voxel. For complex objects composed of multiple sub-components, the volume of each sub-component can be calculated separately and then summed to obtain the overall volumetric feature.

[0099] The acquisition of the surface area features includes: obtaining the surface area data of the historical object from the three-dimensional structural data to obtain the surface area features.

[0100] Surface area features can include the total area of ​​the outer surface of a historical object or a specific surface, and are an important factor affecting the cost of surface treatment processes, coating material consumption, and heat exchange efficiency. Specifically, surface mesh data can be directly read from a 3D model file and the sum of the areas of each triangular facet can be calculated, or the continuous surface can be discretized and summed using numerical integration methods. For objects containing internal cavities or complex surfaces, the outer surface area and inner surface area can be calculated separately, or a specific set of surfaces can be selected for statistical analysis based on application requirements.

[0101] The acquisition of the compact features includes: acquiring the volume data and surface area data, and generating compact features that characterize the spatial composition of the historical objects based on the volume data and surface area data.

[0102] The compactness feature can be used to measure the space utilization efficiency and geometric regularity of a historical object, reflecting the relative relationship between the object's volume and surface area. Specifically, the compactness feature can include compactness, calculated as (π^(1 / 3)*(6*volume)^(2 / 3)) / surface area. A compactness value closer to 1 indicates that the object's shape is closer to an ideal sphere, resulting in higher space utilization efficiency. A smaller compactness value indicates a more irregular or flatter shape, a larger surface area to volume ratio, which may lead to increased processing difficulty or reduced material utilization. The compactness feature can also include other variations, such as the surface area to volume ratio and the ratio of equivalent sphere diameter to feature length, to quantify the geometric regularity of an object from different perspectives.

[0103] In this embodiment, by clearly defining and obtaining the refined volumetric features, surface area features, and compact features in geometric features, a complete mapping path from basic three-dimensional structural data to quantifiable geometric indicators is established. Volumetric features are directly related to material consumption benchmarks, surface area features affect the costs related to surface treatment and thermal management, and compact features comprehensively reflect the space utilization efficiency and manufacturing complexity of an object. These three complement each other and together constitute the basic dimensions for evaluating the geometric attributes of historical objects. This hierarchical and progressive feature extraction strategy not only ensures the accuracy and reproducibility of feature calculations but also provides standardized input elements for the subsequent construction of feature vectors, enabling a unified measurement basis for comparing the geometric similarity between different historical objects.

[0104] In one embodiment, the topological features include at least one of the following: face features, edge features, vertex features, and topological structure features;

[0105] The acquisition of the surface features includes: traversing the three-dimensional structural data, determining the number of surfaces contained in the historical object, and obtaining the surface features.

[0106] The surface features can include the total number of surface units of a historical object, reflecting the fineness and geometric complexity of the object's surface segmentation. Specifically, the count can be performed by traversing all surface units from the mesh data or boundary representation structure of the 3D model file; for polygonal mesh models, the surface feature is the number of polygonal patches; for parametric surface models, the surface can be discretized into an approximate polygonal mesh before counting, or the number of surface units can be directly calculated based on the topological boundaries of the surface patches. The order of magnitude of the surface features directly affects the computational cost of subsequent feature extraction and is also related to the visual fineness and manufacturing precision requirements of the object.

[0107] The acquisition of the edge features includes: traversing the three-dimensional structure data, determining the number of edges contained in the historical object, and obtaining the edge features.

[0108] Edge features can include the total number of edges or edge units in a historical object, used to characterize the complexity and topological characteristics of the object's outline. The specific statistical method varies depending on the model type: for polygonal mesh models, the edge feature is the total number of shared edges between polygonal faces, including boundary edges and internal edges; for parametric curve models, the number can be calculated by discretizing the curve into approximate line segments or by directly calculating the number of edge units based on the topological connectivity of the curve segments.

[0109] The acquisition of point features includes: traversing the three-dimensional structural data, determining the number of points contained in the historical object, and obtaining point features.

[0110] Point features can include the total number of vertices or control points in a historical object, and are the most basic unit for describing the degree of geometric discretization of an object. For polygonal mesh models, point features are the total number of vertices in each polygonal facet; for point cloud data, point features are the number of sampling points in the point cloud; for parametric surface models, point features can include the number of vertices in the control mesh or the number of sampling points on the surface, etc.

[0111] The acquisition of the topological features includes: acquiring the number of faces, edges, and vertices of the historical object, and generating the topological features of the historical object based on the number of faces, edges, and vertices.

[0112] Topological features can include calculating topological invariants based on the number of faces, edges, and vertices to characterize the topological structure of an object. Specifically, the Euler characteristic number can be calculated as a core indicator of topological structure features: Euler characteristic number = number of vertices - number of edges + number of faces. For closed 3D manifolds, the Euler characteristic number has a definite relationship with the object's genus number, reflecting the number of holes on the object's surface; for example, the Euler characteristic number of a sphere is 2, and that of a torus is 0. Furthermore, topological features can also include more refined topological invariants such as the number of connected components, the number of boundary loops, and the Betti number sequence, used to distinguish objects with the same Euler characteristic number but different topological structures; or constructing topological representations of objects such as adjacency graphs or dual graphs, and quantifying the complexity and connectivity of the topological structure using graph theory indicators such as average degree, clustering coefficient, and diameter.

[0113] In this embodiment, a complete index system describing the surface composition and structural complexity of historical objects is constructed through the systematic extraction of face features, edge features, vertex features, and topological features. Face features reflect the degree of geometric discretization and surface complexity of the object: more faces generally mean higher model accuracy or more complex surfaces, but data processing and manufacturing difficulties may also increase accordingly. Edge features characterize the complexity of the object's contour and the number of edges, and are closely related to processing path planning and edge processing techniques. Vertex features, as the basic unit of geometric description, directly affect the model's storage requirements and computational efficiency. Topological features, based on the quantitative relationships of faces, edges, and vertices, further reveal the overall structural attributes and connectivity patterns of the object, providing a solid topological foundation for subsequent similarity comparisons and structural classification.

[0114] In one embodiment, the shape feature includes at least one of the following: a spherical feature, a filling feature, and an elongation feature;

[0115] The acquisition of the spherical feature includes: acquiring the volume of the historical object and the volume of the corresponding minimum circumscribed sphere, and generating the spherical feature of the historical object based on the volume of the historical object and the volume of the minimum circumscribed sphere.

[0116] The sphericity feature can be the ratio of the volume of the historical object to the volume of the minimum bounding sphere, used to quantify how close the object's shape is to an ideal sphere. Specifically, the sphericity feature can be defined as sphericity = (π^(1 / 3)*(6*volume)) / surface area; the closer the sphericity value is to 1, the closer the object's shape is to a sphere, and the more uniform it is in terms of force distribution and fluid resistance; the smaller the sphericity value, the flatter or more irregular the object's shape, and the more likely there is a significant difference between the major and minor axes. The volume of the minimum bounding sphere can be obtained by solving the minimum bounding sphere algorithm for a 3D point set.

[0117] The acquisition of the filling feature includes: acquiring the volume of the historical object and the volume of the corresponding minimum bounding cube, and generating the filling feature of the historical object based on the volume of the historical object and the volume of the minimum bounding cube;

[0118] The filling feature can include the ratio of the volume of the historical object to the volume of the minimum bounding cuboid, used to measure the space-filling efficiency of the object within a regular container. Specifically, the filling feature can be defined as: Fill Rate = Minimum Bounding Cuboid Volume / Historical Object Volume. The closer the filling rate is to 1, the closer the object's geometry is to a regular cuboid; the smaller the filling rate, the more irregular the object's shape or the presence of numerous gaps. The volume of the minimum bounding cuboid can be obtained by calculating the maximum span of the 3D point set in three orthogonal directions (i.e., the product of length, width, and height), or by using principal component analysis to determine the object's principal axis direction and then calculating the minimum bounding cuboid volume along the principal axis direction.

[0119] The acquisition of the elongation feature includes: acquiring the length of the historical object in three principal axis directions, and generating the elongation feature of the historical object based on the correspondence between the lengths in the three principal axis directions.

[0120] The elongation characteristics can include the elongation rate and flatness of historical objects; the elongation rate can be defined as the ratio of the longest principal axis length to the shortest principal axis length, used to characterize the extent of the object's extension along a certain direction; the flatness can be defined as (longest principal axis length * middle principal axis length) / (shortest principal axis length). 2 This describes the relative compression of an object across different dimensions. Specifically, the three principal axes can be obtained by performing principal component analysis on the three-dimensional structural data, corresponding to the three eigenvector directions of the covariance matrix. The length of each principal axis corresponds to the square root of the eigenvalue or the projection span of the three-dimensional point set in that direction.

[0121] In this embodiment, quantitative indicators describing the overall morphology and spatial extension characteristics of historical objects are established by extracting spherical features, filling features, and elongation features. Specifically, spherical features measure the object's proximity to a sphere, reflecting its isotropy and symmetry; filling features assess the object's space utilization within its smallest circumscribed cuboid, revealing its regularity and packaging efficiency; and elongation features characterize the object's extension ratio in different directions, used to distinguish typical forms such as flat, rod-shaped, or balanced shapes. These three types of shape features, along with geometric and topological features, complement each other, forming a multi-level descriptive framework from local details to overall outline, providing a more comprehensive morphological basis for similarity comparisons based on three-dimensional structures.

[0122] In one embodiment, obtaining the feature vector of the query object and indexing the historical objects corresponding to the query object based on the target database includes:

[0123] The feature vector of the query object is compared with multiple historical feature vectors stored in the target database to determine the similarity relationship between the feature vector of the query object and each of the historical feature vectors.

[0124] Based on the similarity relationship, target historical objects that match the feature vector of the query object are determined from the target database.

[0125] In one embodiment, logarithmic similarity can be used to measure basic geometric features. Specifically, the values ​​of the query object and historical objects in features such as volume and surface area are obtained, the square of the difference between their logarithmic values ​​is calculated, and then the difference is converted into a similarity value using a Gaussian kernel function. For size features, a combination of proportional similarity and size proportion consistency can be used. The proportional similarity of each dimension (length, width, and height) is calculated separately and averaged to obtain the size proportion similarity; the three dimensions are then used to form a vector, and the cosine similarity between the size vectors of the query object and historical objects is calculated to measure the consistency of the three-dimensional size proportions; finally, the two are weighted and summed to obtain the comprehensive size feature similarity. For topological features, a combination of proportional similarity and Euler feature similarity can be used. The proportional similarity of topological features such as the number of faces, edges, and vertices is calculated separately and averaged to obtain the topological proportional similarity; then, the Euler characteristic of the object is calculated according to Euler's formula, and the Euler feature similarity is calculated based on the difference between the two Euler characteristic values; finally, the topological proportional similarity and Euler feature similarity are weighted and summed to obtain the comprehensive topological feature similarity. For shape descriptor features, similarity based on the ratio of absolute differences can be used for measurement. Let the feature value of the query object be 'a' and the feature value of the historical object be 'b'. The similarity is defined as 1 minus the absolute difference between 'a' and 'b' divided by the maximum of their absolute values, with a very small constant introduced in the denominator to avoid division by zero. A larger value indicates that the two objects are closer in that shape feature, intuitively reflecting the relative differences in feature values. For rotational features, cosine similarity can be used for measurement. Treating the moments of inertia along the three principal axes as three-dimensional vectors, the cosine similarity between the moment of inertia vectors of the query object and the historical object is calculated. The closer this value is to 1, the more consistent the distribution patterns of the two objects along the principal axes of inertia. For contour features, a similar absolute difference ratio similarity can be used, i.e., 1 minus the absolute difference between the two feature values ​​divided by the maximum of their absolute values. The closer this similarity value is to 1, the more consistent the two objects are in contour features such as convexity and solidity, allowing for a quick assessment of the fullness and convexity differences of the object's outer contour. For symmetrical features, since the mirror symmetry metric is normalized to between 0 and 1, absolute difference similarity can be used for calculation. That is, the similarity equals 1 minus the absolute difference between the two feature values. This method intuitively reflects the closeness of symmetry; a larger value indicates greater similarity. For undulating features, the absolute difference proportional similarity is also used for measurement, which is 1 minus the absolute difference between the two feature values ​​divided by the maximum of their absolute values. This similarity effectively measures the similarity of the surface undulations of an object, providing a quantitative basis for judging whether the surface complexity is similar.

[0126] In one exemplary embodiment, after obtaining the similarity of each feature, the similarity of each feature can be weighted and fused to obtain the comprehensive similarity between the query object and historical objects. Specifically, corresponding weight coefficients can be assigned to geometric features, topological features, shape features, etc., according to the importance of different feature categories to the object recognition and matching task. For example, for small, simple parts, the weight of size and basic geometry can be increased, while the weight of topological and complex features can be decreased; for complex parts, the weight of topological and shape features can be increased, while the weight of size features can be decreased; for large parts, the weight of basic geometry and moment of inertia can be increased, while the weight of curvature and symmetry can be decreased; for flat parts, the weight of symmetry and shape descriptors can be increased, etc.

[0127] In this embodiment, the feature vector of the query object is obtained and compared with multiple historical feature vectors stored in the target database. After determining the similarity relationship, the historical object that best matches the query object is selected from the database. Through this feature vector comparison and similarity calculation method, accurate quantitative matching of historical objects is achieved, improving the accuracy of resource exchange information prediction, while reducing reliance on human experience and subjective judgment errors.

[0128] In one embodiment, before comparing the feature vector of the query object with multiple historical feature vectors stored in the target database, the method further includes:

[0129] Retrieve the volume and length data of the queried object.

[0130] Based on the volume and length data of the queried object, historical feature vectors with volume or size differences less than a preset threshold are filtered from the target database to obtain candidate feature vectors.

[0131] In one exemplary embodiment, pre-screening can be performed before comparing the feature vector of the query object with the historical feature vectors aggregated in the target database. This pre-screening uses the two basic geometric attributes of volume and length to quickly filter out historical objects that differ significantly in overall size from the query object, thereby narrowing the comparison range and reducing the computational overhead of subsequent fine-grained feature comparison. In another exemplary embodiment, pre-screening may further include filtering based on volume, size ratio, topology, etc., to further exclude obviously mismatched objects.

[0132] In this embodiment, by performing preliminary screening based on the volume and length data of the query object before comparing the feature vector of the query object with all historical feature vectors in the target database, the comparison range can be effectively narrowed, the computational complexity reduced, and the retrieval efficiency improved.

[0133] In one embodiment, the feature data further includes at least one of the following: rotational features, contour features, symmetry features, and undulation features;

[0134] The acquisition of the rotation features includes: acquiring the rotation parameters of the historical object, and generating the rotation features of the historical object based on the rotation parameters.

[0135] The rotational features can include principal moments of inertia calculated based on the inertia tensor and their distribution characteristics, used to characterize the inertial properties of the object's mass distribution relative to the rotation axes. Specifically, the three-dimensional inertia tensor matrix of the historical object can be calculated, which describes the distribution of the object's mass relative to each axis of the centroid coordinate system; eigenvalue decomposition of the inertia tensor yields three principal moments of inertia, corresponding to the magnitude of the object's inertia when rotating around the three principal axes. The rotational features can include the numerical magnitudes of the three principal moments of inertia, the ratios between the principal moments of inertia, and the normalized distribution vectors of the principal moments of inertia.

[0136] The acquisition of the outline features includes: acquiring the volume of the minimum convex hull of the historical object and the volume of the historical object, and generating the outline features of the historical object based on the volume of the minimum convex hull and the volume of the historical object itself;

[0137] The external contour features can include metrics such as convexity and solidity, used to describe the fullness of the object's outer contour and the compactness of its internal structure. Specifically, convexity can be defined as the ratio of the historical object volume to its minimum convex hull volume. A convexity value closer to 1 indicates a more convex surface with fewer depressions; a smaller convexity value indicates more depressions or holes on the object's surface. Solidity can be defined as the ratio of the historical object volume to its minimum bounding sphere volume, used to measure the object's filling density within the minimum bounding sphere. A higher solidity value indicates a more compact object, while a lower value indicates a looser shape or more suspended structures. The volume of the minimum convex hull can be calculated using convex hull algorithms such as the fast convex hull algorithm and the incremental convex hull algorithm.

[0138] The acquisition of the symmetry features includes: detecting the symmetry relationship of the historical object in three orthogonal directions, and generating the symmetry features of the historical object based on the symmetry relationship;

[0139] Symmetry features can include metrics such as mirror symmetry and rotational symmetry, used to quantify the degree of symmetry in an object's geometric structure. Specifically, mirror symmetry can be obtained by calculating the degree of overlap between the object and the original object after reflection transformation about a plane; a higher overlap ratio indicates stronger symmetry. This can be calculated using three orthogonal planes (such as the XY plane, YZ plane, and ZX plane) or candidate symmetry planes determined by principal component analysis, yielding a multi-dimensional symmetry measure. Rotational symmetry can be evaluated by analyzing the degree of self-coincidence of an object after rotating it around a certain axis; it is suitable for objects with axial or rotational symmetry characteristics. Symmetry features can include the symmetry values ​​of each symmetry plane, the maximum symmetry value, and the symmetry distribution entropy, used to distinguish highly symmetrical objects from irregularly shaped objects.

[0140] The acquisition of the undulation features includes: acquiring the curvature data of the surface of the historical object and the distribution of the curvature data, and generating the undulation features of the historical object based on the curvature data and the distribution of the curvature data.

[0141] The undulation features can include the statistical distribution characteristics and spatial variation properties of surface curvature, used to characterize the smoothness and local undulation of an object's surface. Specifically, the Gaussian curvature and mean curvature of each vertex or facet of a historical object's surface can be calculated, and then the distribution parameters of curvature such as mean, variance, maximum, minimum, skewness, and kurtosis can be statistically analyzed. Heatmap distribution of curvature, rate of change of curvature gradient, and the proportion of regions with curvature exceeding a specific threshold can also be calculated. The undulation features can also include surface roughness indices, such as arithmetic mean roughness and root mean square roughness, obtained by quantifying the deviation between the actual surface and an ideal smooth surface. For triangular mesh models, local curvature can be estimated by the change in the angle between the normal vectors of adjacent facets, or the curvature tensor can be calculated using discrete differential geometry methods.

[0142] In this embodiment, the feature description dimensions of historical objects are further enriched by extracting rotational features, contour features, symmetry features, and undulation features. Rotational features are obtained based on rotational parameters such as the principal axis moment of inertia, used to quantify the inertial distribution characteristics of the object during rotational motion. Contour features reflect the fullness and concavity ratio of the object's outer contour by calculating the ratio of the minimum convex hull volume to the object's own volume. Symmetry features quantify the strength of the object's symmetry by detecting the mirror symmetry relationship in three orthogonal directions. Undulation features characterize the degree of local changes on the object's surface based on bending data and their distribution, such as the mean and standard deviation of surface curvature. By introducing multi-dimensional deep features, a comprehensive digital description of the object's physical properties and morphological features is achieved, thereby improving the discriminative power and robustness of subsequent similarity matching and reducing structural misjudgments caused by single features.

[0143] In one embodiment, the data processing method provided in this application can be applied to an intelligent quotation system in the machining industry, such as... Figure 3 As shown, the specific steps may include the following:

[0144] Step S300: Obtain historical part data. Historical part 3D drawings and their corresponding historical pricing data can be imported into the system. The 3D drawings can be in STP or STEP format; the pricing data can include pricing-related information such as process difficulty, working hours, material costs, market prices, and transaction prices.

[0145] Step S302: Extract 3D structural features. The 3D drawing file for each historical part is analyzed to extract eight types of feature data. Specifically, geometric features are obtained by calculating volume, surface area, and compactness; dimensional features are obtained by acquiring the length, width, and height of the bounding box in the drawing file; topological features are obtained by counting the number of faces, edges, and vertices; shape descriptors are obtained by calculating sphericity, rectangularity, elongation, and flatness; moment of inertia features are obtained by extracting the moments of inertia of the three principal axes; convex hull features are obtained by calculating convexity and solidity; symmetry features are obtained by detecting mirror symmetry in three orthogonal directions; and curvature features are obtained by calculating the mean and standard deviation of curvature.

[0146] Step S304: Construct and store the feature vector. The eight types of features are combined into a high-dimensional feature vector, and different features are processed using appropriate normalization methods to ensure vertical comparability across dimensions. Geometric features are normalized using logarithmic ratio similarity, dimensional and topological features using proportional similarity, moment of inertia features using cosine similarity, and symmetry features using absolute difference similarity. The normalized feature vector, along with its corresponding feature hash value, is stored in the database and associated with the identifiers of historical parts, forming a searchable feature index.

[0147] Step S306: Obtain the features of the query part. When a quote is needed for a new part, upload its 3D drawing file, extract the eight features of the query part, and generate a normalized feature vector.

[0148] Step S308: Matching similar historical parts. Historical feature vectors obtained from the database can be filtered based on the volume and dimensions of the queried part to quickly eliminate parts with significant differences, thereby selecting candidate historical parts. The similarity between the queried part and the candidate historical parts in each dimension of features is calculated one by one. The similarities in each dimension are then weighted and summed to obtain a comprehensive similarity score. The weights of the similarity scores in each dimension can be dynamically adjusted based on the part type, etc.

[0149] Step S310: Generate reference quotation information. Based on the matching results, if the similarity is greater than a preset threshold, the quotation information of the matched parts can be directly applied, and fine-tuned according to the material and process data of the current parts to obtain reference quotation information. If the similarity is less than the preset threshold, reference quotation information can be directly predicted based on the volume, density, process difficulty, material, and labor time data of the queried parts.

[0150] Step S312: Output the quotation results. Display the generated quotation information to the user, including material costs, processing costs, process difficulty, working hours, market quotations, transaction prices, and other quotation-related information, and save the results to the database.

[0151] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0152] Based on the same inventive concept, this application also provides a data processing apparatus for implementing the data processing method described above. The solution provided by this apparatus is similar to the implementation scheme described in the above method; therefore, the specific limitations in one or more data processing apparatus embodiments provided below can be found in the limitations of the data processing method described above, and will not be repeated here.

[0153] In one embodiment, such as Figure 4 As shown, a data processing device 400 is provided, including: a data acquisition module 401, a feature extraction module 403, a data storage module 405, a data indexing module 407, and a data generation module 409, wherein:

[0154] The data acquisition module is used to acquire the three-dimensional structural data and resource exchange information of historical objects;

[0155] The feature extraction module is used to extract features from the three-dimensional structural data to obtain feature data of the historical object; wherein, the feature data includes at least one of the following: geometric features, topological features, and shape features;

[0156] The feature extraction module is also used to determine the feature vector of the historical object based on the feature data;

[0157] The data storage module is used to associate the feature vector with the identifier of the historical object and store it in the target database;

[0158] The data indexing module is used to obtain the feature vector of the query object, and match the target historical object that matches the feature vector of the query object from the target database based on the similarity relationship between the feature vector of the query object and the feature vector of each historical object.

[0159] The data generation module is used to generate reference resource exchange information for the queried object based on the resource exchange information of the target historical object; the resource exchange information is used to characterize the quantity of exchangeable resources of the object.

[0160] In one embodiment, the geometric features include at least one of the following: volume features, surface area features, and compactness features; the feature extraction module is further configured to:

[0161] Volume data of the historical object is obtained from the three-dimensional structural data to obtain volume features;

[0162] The surface area data of the historical object is obtained from the three-dimensional structural data to obtain surface area features;

[0163] The volume data and surface area data are acquired, and a compact feature characterizing the spatial composition of the historical object is generated based on the volume data and surface area data.

[0164] In one embodiment, the topological features include at least one of the following: face features, edge features, vertex features, and topological structure features; the feature extraction module is further configured to:

[0165] Traverse the three-dimensional structural data to determine the number of faces contained in the historical object and obtain the face features;

[0166] Traverse the three-dimensional structure data to determine the number of edges contained in the historical object and obtain the edge features;

[0167] Traverse the three-dimensional structural data to determine the number of points contained in the historical object and obtain point features;

[0168] Obtain the number of faces, edges, and vertices of the historical object, and generate the topological features of the historical object based on the number of faces, edges, and vertices.

[0169] In one embodiment, the shape feature includes at least one of the following: a spherical feature, a filled feature, and an elongated feature; the feature extraction module is further configured to:

[0170] Obtain the volume of the historical object and the volume of the corresponding minimum circumscribed sphere, and generate the spherical feature of the historical object based on the volume of the historical object and the volume of the minimum circumscribed sphere;

[0171] Obtain the volume of the historical object and the volume of the corresponding minimum bounding box, and generate the filling feature of the historical object based on the volume of the historical object and the volume of the minimum bounding box;

[0172] The lengths of the historical object in the three principal axis directions are obtained, and the elongation features of the historical object are generated according to the correspondence between the lengths in the three principal axis directions.

[0173] In one embodiment, the data indexing module is further configured to:

[0174] The feature vector of the query object is compared with multiple historical feature vectors stored in the target database to determine the similarity relationship between the feature vector of the query object and each of the historical feature vectors.

[0175] Based on the similarity relationship, target historical objects that match the feature vector of the query object are determined from the target database.

[0176] In one embodiment, the apparatus further includes a data filtering module for:

[0177] Retrieve the volume and length data of the queried object;

[0178] Based on the volume and length data of the queried object, historical feature vectors with volume or size differences less than a preset threshold are filtered from the target database to obtain candidate feature vectors.

[0179] In one embodiment, the feature data further includes at least one of the following: rotational features, contour features, symmetry features, and undulation features; the feature extraction module is further configured to:

[0180] Obtain the rotation parameters of the historical object, and generate the rotation features of the historical object based on the rotation parameters;

[0181] Obtain the volume of the minimum convex hull of the historical object and the volume of the historical object itself, and generate the outline feature of the historical object based on the volume of the minimum convex hull and the volume of the historical object itself;

[0182] The symmetry relationship of the historical object in three orthogonal directions is detected, and the symmetry feature of the historical object is generated based on the symmetry relationship;

[0183] Obtain the curvature data of the surface of the historical object and the distribution of the curvature data, and generate the undulation features of the historical object based on the curvature data and the distribution of the curvature data.

[0184] Each module in the aforementioned data processing device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the operations corresponding to each module.

[0185] In one embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 5 As shown, the computer device includes a processor, memory, input / output interfaces, a communication interface, a display unit, and an input device. The processor, memory, and input / output interfaces are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interfaces. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The input / output interfaces are used for exchanging information between the processor and external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, NFC (Near Field Communication), or other technologies. When the computer program is executed by the processor, it implements a data processing method. The display unit is used to form a visually visible image and can be a display screen, a projection device, or a virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The input device of the computer device can be a touch layer covering the display screen, or buttons, trackballs, or touchpads set on the casing of the computer device, or external keyboards, touchpads, or mice, etc.

[0186] Those skilled in the art will understand that Figure 5 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0187] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of related data must comply with the relevant laws, regulations and standards of the relevant countries and regions.

[0188] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0189] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0190] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A data processing method, characterized in that, The method includes: Acquire the 3D structural data and resource exchange information of historical objects; Feature extraction is performed on the three-dimensional structural data to obtain the feature data of the historical object; wherein, the feature data includes at least one of the following: geometric features, topological features, and shape features; Based on the feature data, the feature vector of the historical object is determined; The feature vector is associated with the identifier of the historical object and stored in the target database; Obtain the feature vector of the query object, and based on the similarity relationship between the feature vector of the query object and the feature vectors of each historical object, match the target historical object that matches the feature vector of the query object from the target database; Based on the resource exchange information of the target historical object, reference resource exchange information of the queried object is generated; the resource exchange information is used to characterize the quantity of exchangeable resources of the object.

2. The method according to claim 1, characterized in that, The geometric features include at least one of the following: volume features, surface area features, and compact features; The acquisition of volume features includes: obtaining volume data of the historical object from the three-dimensional structural data to obtain volume features; The acquisition of the surface area features includes: obtaining the surface area data of the historical object from the three-dimensional structural data to obtain the surface area features; The acquisition of the compact features includes: acquiring the volume data and surface area data, and generating compact features that characterize the spatial composition of the historical objects based on the volume data and surface area data.

3. The method according to claim 1, characterized in that, The topological features include at least one of the following: face features, edge features, vertex features, and topological structure features; The acquisition of the surface features includes: traversing the three-dimensional structural data, determining the number of surfaces contained in the historical object, and obtaining the surface features; The acquisition of the edge features includes: traversing the three-dimensional structure data, determining the number of edges contained in the historical object, and obtaining the edge features; The acquisition of point features includes: traversing the three-dimensional structural data, determining the number of points contained in the historical object, and obtaining point features; The acquisition of the topological features includes: acquiring the number of faces, edges, and vertices of the historical object, and generating the topological features of the historical object based on the number of faces, edges, and vertices.

4. The method according to claim 1, characterized in that, The shape feature includes at least one of the following: spherical feature, filling feature, and elongation feature; The acquisition of the spherical feature includes: acquiring the volume of the historical object and the volume of the corresponding minimum circumscribed sphere, and generating the spherical feature of the historical object based on the volume of the historical object and the volume of the minimum circumscribed sphere; The acquisition of the filling feature includes: acquiring the volume of the historical object and the volume of the corresponding minimum bounding cube, and generating the filling feature of the historical object based on the volume of the historical object and the volume of the minimum bounding cube; The acquisition of the elongation feature includes: acquiring the length of the historical object in three principal axis directions, and generating the elongation feature of the historical object based on the correspondence between the lengths in the three principal axis directions.

5. The method according to claim 1, characterized in that, The step of matching target historical objects that match the feature vector of the query object from the target database based on the similarity relationship between the feature vector of the query object and the feature vectors of each historical object includes: The feature vector of the query object is compared with multiple historical feature vectors stored in the target database to determine the similarity relationship between the feature vector of the query object and each of the historical feature vectors. Based on the similarity relationship, target historical objects that match the feature vector of the query object are determined from the target database.

6. The method according to claim 5, characterized in that, Before comparing the feature vector of the query object with multiple historical feature vectors stored in the target database, the process also includes: Retrieve the volume and length data of the queried object; Based on the volume and length data of the queried object, historical feature vectors with volume or size differences less than a preset threshold are filtered from the target database to obtain candidate feature vectors.

7. The method according to claim 1, characterized in that, The feature data also includes at least one of the following: rotation features, outline features, symmetry features, and undulation features; The acquisition of the rotation features includes: acquiring the rotation parameters of the historical object, and generating the rotation features of the historical object based on the rotation parameters; The acquisition of the outline features includes: acquiring the volume of the minimum convex hull of the historical object and the volume of the historical object, and generating the outline features of the historical object based on the volume of the minimum convex hull and the volume of the historical object itself; The acquisition of the symmetry features includes: detecting the symmetry relationship of the historical object in three orthogonal directions, and generating the symmetry features of the historical object based on the symmetry relationship; The acquisition of the undulation features includes: acquiring the curvature data of the surface of the historical object and the distribution of the curvature data, and generating the undulation features of the historical object based on the curvature data and the distribution of the curvature data.

8. A data processing apparatus, characterized in that, The device includes: The data acquisition module is used to acquire the three-dimensional structural data and resource exchange information of historical objects; The feature extraction module is used to extract features from the three-dimensional structural data to obtain feature data of the historical object; wherein, the feature data includes at least one of the following: geometric features, topological features, and shape features; The feature extraction module is also used to determine the feature vector of the historical object based on the feature data; The data storage module is used to associate the feature vector with the identifier of the historical object and store it in the target database; The data indexing module is used to obtain the feature vector of the query object, and match the target historical object that matches the feature vector of the query object from the target database based on the similarity relationship between the feature vector of the query object and the feature vector of each historical object. The data generation module is used to generate reference resource exchange information for the queried object based on the resource exchange information of the target historical object; the resource exchange information is used to characterize the quantity of exchangeable resources of the object.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.

Images