Three-dimensional model processing method and system training and application method, equipment and medium

Abstract

The invention relates to the field of mechanical design and manufacturing, and provides a three-dimensional model processing method, a training and application method of a system, equipment and a medium, and a target boundary representation three-dimensional model is obtained based on a preset complex manufacturing feature model and a boundary representation three-dimensional model. The complex manufacturing feature model is added to the boundary representation three-dimensional model containing the simple structure manufacturing feature model, so that the boundary representation three-dimensional model containing the simple structure manufacturing feature model and the complex structure manufacturing feature model is synthesized. According to the three-dimensional model processing scheme provided by the invention, the complex manufacturing feature model is added in the boundary representation three-dimensional model, so that when the boundary representation three-dimensional model obtained by the processing scheme is used for training an image classification system, the image classification efficiency is improved; the generalization ability of the image classification system and the classification accuracy of the manufacturing feature model in the boundary representation three-dimensional model can be improved, so that the production efficiency and quality of process products are improved.

Description

Technical Field

[0001] The present application relates to the field of mechanical design and manufacturing, and in particular to a three-dimensional model processing method, system training and application method, equipment, and media. Background Art

[0002] In the field of mechanical design and manufacturing, the boundary representation 3D model within the CAD model is an important vehicle for process product design. This boundary representation 3D model contains not only geometric information but also design intent and manufacturing feature models. Manufacturing feature recognition (MFR) is a key technology that connects computer-aided design (CAD), computer-aided engineering (CAE), computer-aided process planning (CAPP), and computer-aided manufacturing (CAM) systems. It reanalyzes the boundary representation 3D model from a manufacturing perspective, accurately identifying and classifying the manufacturing feature models within the boundary representation 3D model. This allows for accurate production process planning based on the manufacturing feature models, ultimately resulting in high-quality process products. For example, manufacturing features such as holes, slots, steps, and chamfers can be classified to produce high-quality manufacturing process products. Complex manufacturing feature models (such as clips, screws, bosses, and ribs) are very common in mechanical design, but these feature models typically have complex geometric shapes and topological structures and often intersect or overlap with other features. Existing manufacturing feature model classification methods face various challenges in classifying these complex manufacturing feature models.

[0003] The inventors have discovered that deep learning methods have made significant progress in the field of feature recognition in recent years, particularly deep learning methods based on point clouds, voxels, polygonal meshes, and boundary representation 3D models (also known as B-rep models). However, due to the lack of a large amount of valid data on boundary representation 3D models, existing deep learning-based classification methods can only train image classification systems using a small number of effective and simple manufacturing feature models. This results in low generalization capabilities and low classification accuracy for the trained image classification systems. In other words, when using a trained image classification system to classify complex manufacturing feature models in boundary representation 3D models, inaccurate classification results are highly likely to be obtained.

[0004] In summary, the existing deep learning-based methods for classifying manufacturing feature models in boundary representation three-dimensional models lack a large number of effective and complex manufacturing feature models, resulting in low generalization ability of the trained image classification system and low classification accuracy of the manufacturing feature models. As a result, when the trained image classification system is used to classify the complex manufacturing feature models in the boundary representation three-dimensional model, inaccurate classification results are likely to be obtained, resulting in a decrease in the production efficiency and quality of the process products. Summary of the Invention

[0005] This application provides a three-dimensional model processing method, system training and application method, equipment, and medium to solve the problem of low classification accuracy when classifying manufacturing feature models in boundary representation three-dimensional models based on existing deep learning methods.

[0006] A first aspect of the present application provides a three-dimensional model processing method, system training and application method, device, and medium. The method is used to process a boundary representation three-dimensional model, and the method includes:

[0007] Obtaining a boundary representation three-dimensional model;

[0008] Based on the boundary representation three-dimensional model and the preset complex manufacturing feature model, the target boundary representation three-dimensional model is calculated.

[0009] In some embodiments of the present application, the boundary representation three-dimensional model includes: a base model and a manufacturing feature model, and the step of calculating the target boundary representation three-dimensional model based on the boundary representation three-dimensional model and the preset complex manufacturing feature model includes:

[0010] The complex manufacturing feature model is moved, and / or copied, and / or rotated, and / or scaled on the base model, and the processed complex manufacturing feature model and the manufacturing feature model are fused with the base model to obtain a target boundary representation three-dimensional model.

[0011] In some embodiments of the present application, the step of moving the complex manufacturing feature model on the base model includes:

[0012] The complex manufacturing feature model is translated on the base model, so that the movement process of the complex manufacturing feature model is completed on the base model.

[0013] In some embodiments of the present application, the step of replicating the complex manufacturing feature model on the base model includes:

[0014] The complex manufacturing feature model is copied on the base model, so that the copying process of the complex manufacturing feature model is completed on the base model.

[0015] In some embodiments of the present application, the step of rotating the complex manufacturing feature model on the base model includes:

[0016] The complex manufacturing feature model is rotated on the base model according to a preset rotation angle, so that the rotation processing of the complex manufacturing feature model is completed on the base model.

[0017] In some embodiments of the present application, the step of scaling the complex manufacturing feature model on the base model includes:

[0018] The complex manufacturing feature model is scaled on the base model, so that the scaling process of the complex manufacturing feature model is completed on the base model.

[0019] A second aspect of the present application provides a training method for an image classification system, wherein the image classification system is used to classify boundary representation surfaces in a boundary representation three-dimensional model, and the method comprises:

[0020] Obtaining a boundary representation three-dimensional model;

[0021] Using the three-dimensional model processing method of the above embodiment, the boundary representation three-dimensional model is processed, and the processed results form a data set;

[0022] The image classification system is trained using the dataset for multiple iterations until convergence.

[0023] A third aspect of the present application provides an application method of an image classification system, the method comprising:

[0024] Acquire a boundary representation three-dimensional model, wherein the boundary representation three-dimensional model includes a boundary representation surface;

[0025] The image classification system trained by the training method of the image classification system of the above embodiment is used to classify the boundary representation surfaces in the boundary representation three-dimensional model.

[0026] The fourth aspect of the present application provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the method described in any one of the first, second, and third aspects of the above embodiments are implemented.

[0027] The fifth aspect of the present application provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the method described in any one of the first, second, and third aspects of the above embodiments is implemented.

[0028] This application has the following beneficial effects:

[0029] The above-mentioned embodiments of the present application are based on a preset complex manufacturing feature model and a boundary representation three-dimensional model to obtain a target boundary representation three-dimensional model. The present application adds a complex manufacturing feature model to the boundary representation three-dimensional model containing a simple structure manufacturing feature model to synthesize a boundary representation three-dimensional model containing simple structure and complex structure manufacturing feature models. The three-dimensional model processing scheme of the present application adds a complex manufacturing feature model to the boundary representation three-dimensional model. Therefore, when the image classification system is trained using the boundary representation three-dimensional model obtained by the processing scheme of the present application, the generalization ability of the image classification system and the classification accuracy of the manufacturing feature model in the boundary representation three-dimensional model can be improved, thereby improving the production efficiency and quality of the process products. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] The drawings herein are incorporated into and constitute a part of the specification. These drawings illustrate embodiments consistent with the present application and, together with the specification, are used to illustrate the technical solutions of the present application.

[0031] Figure 1 This is a flow chart of an embodiment of a method for processing a three-dimensional model provided by the present application;

[0032] Figure 2 This is a first example schematic diagram of a base model and a complex manufacturing feature model provided by this application;

[0033] Figure 3 is a schematic diagram of experimental results of performing a translation operation on the complex manufacturing feature model in the first example provided by the present application;

[0034] Figure 4 is a schematic diagram of experimental results of replicating the complex manufacturing feature model in the first example provided by the present application;

[0035] Figure 5 is a schematic diagram of experimental results of rotating multiple complex manufacturing feature models in the first example provided by the present application;

[0036] Figure 6 is a schematic diagram of experimental results of scaling the complex manufacturing feature model in the first example provided by the present application;

[0037] Figure 7 is a second example schematic diagram of a base model and a complex manufacturing feature model provided by this application;

[0038] Figure 8 is a schematic diagram of experimental results of performing a translation operation on the complex manufacturing feature model in the second example provided by this application;

[0039] Figure 9 is a schematic diagram of experimental results of replicating the complex manufacturing feature model in the second example provided by the present application;

[0040] Figure 10 is a schematic diagram of experimental results of rotating the complex manufacturing feature model in the second example provided by the present application;

[0041] Figure 11 is a schematic diagram of experimental results of scaling the complex manufacturing feature model in the second example provided by the present application;

[0042] Figure 12 This is a flow chart of an embodiment of a training method for an image classification system provided by the present application;

[0043] Figure 13 This is a schematic diagram of an example of the experimental results of data enhancement processing and data annotation provided in this application;

[0044] Figure 14 This is a flow chart of an embodiment of an application method of the image classification system provided by the present application;

[0045] Figure 15 This is a schematic diagram of the framework of an embodiment of an electronic device provided by the present application;

[0046] Figure 16 It is a schematic diagram of a framework of an embodiment of a computer-readable storage medium provided by this application. DETAILED DESCRIPTION

[0047] The following describes the embodiments of the present application in detail with reference to the accompanying drawings.

[0048] In the following description, for the purpose of explanation rather than limitation, specific details such as specific system structures, interfaces, and technologies are provided to facilitate a thorough understanding of the present application.

[0049] The term "and / or" in this article is simply a description of the association relationship of associated objects, indicating that three relationships can exist. For example, A and / or B can represent three situations: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects are in an "or" relationship. In addition, "many" in this article means two or more than two. In addition, the term "at least one" in this article means any combination of at least two of any one or more of a plurality of. For example, including at least one of A, B, and C can mean including any one or more elements selected from the set consisting of A, B, and C.

[0050] As described in the background technology, the existing scheme for classifying manufacturing feature models of boundary representation three-dimensional models based on deep learning methods lacks a large number of effective and complex manufacturing feature models, resulting in low generalization ability of the trained image classification system and low classification accuracy of the manufacturing feature models. As a result, when using the trained image classification system to classify complex manufacturing feature models in the boundary representation three-dimensional model, it is very likely to obtain inaccurate classification results.

[0051] To address the above-mentioned issues, the present application proposes a data enhancement scheme for three-dimensional models. In the scheme of the present application, a complex manufacturing feature model is added to a boundary representation three-dimensional model containing a simple structure manufacturing feature model to synthesize a boundary representation three-dimensional model containing both simple and complex structure manufacturing feature models. The three-dimensional model processing scheme of the present application adds a complex manufacturing feature model to the boundary representation three-dimensional model. Therefore, when an image classification system is trained using the boundary representation three-dimensional model obtained by the processing scheme of the present application, the generalization ability of the image classification system and the classification accuracy of the manufacturing feature models in the boundary representation three-dimensional model can be improved.

[0052] Before introducing the specific embodiments of the present application, we first briefly introduce some basic knowledge involved in the present application.

[0053] Boundary representation of three-dimensional models is a commonly used three-dimensional entity representation method in CAD systems. Its core is to describe the boundaries of CAD models through topological entities (such as vertices, edges, faces, rings, shells, etc.) and geometric entities (such as curves and surfaces).

[0054] The whole process of producing a mechanical part:

[0055] Design a part model containing holes, slots and chamfers in CAD software;

[0056] CAE software performed stress analysis on the CAD designed part model and found that the edge stress of the hole was too high;

[0057] Plan the processing technology (such as drilling, milling, chamfering) of the part model in CAPP software based on the analysis results of the part model by CAE software;

[0058] CAM software generates NC codes based on process planning and controls the machine tool to complete the processing of the part model.

[0059] Through MFR, seamless data transfer between CAD, CAE, CAPP, and CAM systems is achieved, reducing manual intervention and improving product production efficiency and quality.

[0060] The present application is described below with reference to the accompanying drawings and specific embodiments.

[0061] According to one embodiment of the present application, Figure 1 As shown, the present application provides a method for processing a three-dimensional model, the method comprising: S1, obtaining a boundary representation three-dimensional model; S2, calculating a target boundary representation three-dimensional model based on the boundary representation three-dimensional model and a preset complex manufacturing feature model. It should be noted that the boundary representation three-dimensional model of the present application can be a two-dimensional image of a boundary representation three-dimensional model (which is a type of CAD model) derived from CAD software.

[0062] For example, the complex manufacturing feature models of the present application include but are not limited to clips, screws, boss columns, reinforcement ribs, etc.

[0063] In order to understand the present application more clearly, step S2 is described in detail below.

[0064] According to one embodiment of the present application, the boundary representation three-dimensional model includes: a base model and a manufacturing feature model, and step S2 includes: moving, and / or copying, and / or rotating, and / or scaling the complex manufacturing feature model on the base model, and fusing the processed complex manufacturing feature model and the manufacturing feature model with the base model to obtain the target boundary representation three-dimensional model. It should be noted that the operations of processing the complex manufacturing feature model on the base model include but are not limited to moving, and / or copying, and / or rotating, and / or scaling, and other processing methods can also be performed according to data requirements, which are not specifically limited here.

[0065] It can be seen from the above description that the above embodiments of the present application can efficiently generate a target boundary representation three-dimensional model by moving, and / or copying, and / or rotating, and / or scaling the complex manufacturing feature model on the base model, and fusing the processed complex manufacturing feature model with the base model and the manufacturing feature model, thereby increasing the number and diversity of manufacturing feature models on the boundary representation three-dimensional model, thereby improving the generalization ability of the image classification system and the classification accuracy of the manufacturing feature models in the boundary representation three-dimensional model.

[0066] Among them, according to one embodiment of the present application, the complex manufacturing feature model is obtained in the following manner: the base model and multiple complex manufacturing feature models are divided on the boundary representation three-dimensional model containing the complex manufacturing feature model in the CAD library to obtain multiple complex manufacturing feature models; the multiple complex manufacturing feature models are moved, and / or copied, and / or rotated, and / or scaled on the base model respectively.

[0067] Separate the complex base model (base) and the complex manufacturing feature model (such as buckles, boss columns, screws, etc.) in the real CAD data. According to one embodiment of the present application, the steps of obtaining multiple complex manufacturing feature models include: traversing the faces, edges, vertices and other geometric entities in the boundary representation three-dimensional model containing the complex manufacturing feature model, and extracting the geometric properties of the faces and edges based on the faces, edges, vertices and other geometric entities; screening the faces in the boundary representation three-dimensional model containing the complex manufacturing feature model based on preset geometric constraint rules; segmenting the faces, edges and vertices in the complex manufacturing feature model from the complex base model; checking the closedness and manifold properties of the segmented models (base model and complex manufacturing feature model); filling holes and stitching edges of the segmented complex base model (to avoid geometric errors caused by topological breaks) to repair the complex base model and complex manufacturing feature model.

[0068] It can be seen from the above embodiments that the geometric segmentation method of the above embodiments of the present application accurately separates the complex manufacturing feature model from the base model by traversing the boundary representation of geometric entities such as faces, edges, vertices in the three-dimensional model, and screening based on geometric constraint rules, thereby ensuring the geometric accuracy and efficiency of the separation process; at the same time, the segmented model is checked for closure and manifoldness, and holes are filled and edges are stitched, which effectively avoids geometric errors caused by topological breaks, ensures the geometric validity and topological integrity of the base model and the complex manufacturing feature model, and thus improves the reliability and practicality of model segmentation.

[0069] According to one embodiment of the present application, the topological query interface of Open CASCADE Technology (OCCT) (such as TopExp_Explorer, TopoDS_Iterator) is used to traverse the faces, edges, vertices and other entities in the B-rep model to extract geometric properties (for example, surface type, area, normal vector); the BRepAlgoAPI_Cut method (i.e., Boolean difference operation) is used to segment the faces, edges, and vertices in the complex manufacturing feature model from the complex base model; the BRepCheck_Analyzer topological inspection method is used to check the closure and manifold properties of the segmented models (complex base model and complex manufacturing feature model); the BRepBuilderAPI_Sewing repair method is used to fill holes and sew edges of the segmented complex base model (to avoid geometric errors caused by topological breaks) to repair the complex base model and the complex manufacturing feature model.

[0070] It can be seen from the above embodiments that the above embodiments of the present application use the topological query interface of Open CASCADE Technology (OCCT) (such as TopExp_Explorer, TopoDS_Iterator) to traverse the entities such as faces, edges, vertices in the B-rep model and extract geometric attributes, which can efficiently and accurately obtain the geometric information of the model and provide a reliable data basis for subsequent geometric segmentation and feature recognition; at the same time, based on the powerful geometric processing capabilities of OCCT, the efficiency and stability of topological query and attribute extraction are ensured, and the complex manufacturing feature model is accurately segmented from the base model based on the geometric constraint rules, combined with the Boolean difference operation (BRepAlgoAPI_Cut), ensuring the geometric accuracy and efficiency of the segmentation process; at the same time, the topological inspection method (BRepCheck_Analyzer) and the repair method (BRepBuilderAPI_Sewing) are used to check the closure and manifold properties of the segmented model, as well as to fill holes and stitch edges, effectively avoiding geometric errors caused by topological breaks, ensuring the geometric validity and topological integrity of the base model and the complex manufacturing feature model, thereby improving the reliability and practicality of model segmentation.

[0071] According to another embodiment of the present application, the geometric constraint rules are coplanarity, and / or concentricity, and / or tangency, and the coplanarity, and / or concentricity, and / or tangency of the faces in the three-dimensional model representing the boundary containing the complex manufacturing feature model are calculated to screen out the complex manufacturing feature model and the base model. It should be noted that the preset geometric constraint rules include but are not limited to coplanarity, concentricity, and tangency. Among them, coplanarity: calculate the angle between the normal vectors of the two faces (such as less than 5°), and check whether the center of mass distance is within the threshold (such as less than 1mm); concentricity: for cylindrical or circular features, check whether the axes coincide (such as extracting the cylinder axis through BRepAdaptor_Surface and calculating the distance between the axes); tangency: check whether the connecting edges between adjacent faces meet the tangency condition (such as using BRepLProp_SLProps to calculate the tangent vector continuity of the edge).

[0072] As can be seen from the above embodiments, the above embodiments of the present application screen the faces in the boundary representation three-dimensional model containing the complex manufacturing feature model based on the preset geometric constraint rules, which can accurately identify and distinguish the geometric relationship between the base model and the complex manufacturing feature model, and ensure the accuracy and efficiency of the segmentation process. Geometric constraint rules such as coplanarity, concentricity and tangency can effectively capture the typical geometric associations between the complex manufacturing feature model and the base model. Through the screening of geometric constraint rules, it is possible to avoid mis-segmentation or omission of key features, and at the same time provide accurate input for subsequent Boolean difference operations, ensuring the geometric and topological integrity and effectiveness of the segmented base model and the complex manufacturing feature model, thereby improving the reliability and practicality of model segmentation.

[0073] According to one embodiment of the present application, in step S1, the boundary representation 3D model is a boundary representation 3D model used to obtain a complex manufacturing feature model, or another boundary representation 3D model that does not include a complex manufacturing feature model. The specific choice can be determined based on task requirements and is not specifically limited here.

[0074] The following describes a method for moving, copying, rotating, and / or scaling a complex manufacturing feature model on a base model.

[0075] (1) Mobile

[0076] According to one embodiment of the present application, the step of moving the complex manufacturing feature model on the base model includes: translating the complex manufacturing feature model on the base model so as to complete the movement of the complex manufacturing feature model on the base model. According to another embodiment of the present application, when the base model includes multiple boundary representation surfaces, the step of moving the complex manufacturing feature model on the base model includes: translating the complex manufacturing feature model at different positions of any boundary representation surface of the base model so as to complete the movement of the complex manufacturing feature model on the base model.

[0077] From the above description, it can be seen that the above embodiment of the present application translates the complex manufacturing feature model at different positions of the base model. This method helps to reduce the cost of constructing the data set required for training the image classification system, and uses the complex manufacturing feature models at different positions to train the image classification system, which can significantly improve the generalization ability of the image classification system and the classification accuracy of the manufacturing feature model in the boundary representation three-dimensional model.

[0078] Among them, according to one embodiment of the present application, the steps of completing the movement processing of the complex manufacturing feature model on the base model include: randomly selecting a point on the contact surface of the base model, wherein the contact surface of the base model is the intersection surface of the base model and the complex manufacturing feature model; using the BRepBuilderAPl_Transform method to translate along the normal vector direction of the point, and completing the movement processing of the complex manufacturing feature model when the complex manufacturing feature model intersects with the base model.

[0079] It can be seen from the above description that the above embodiment of the present application completes the movement processing of the complex manufacturing feature model when the complex manufacturing feature model intersects with the base model. This method can ensure the geometric topological continuity between the complex manufacturing feature model and the base model after movement, effectively avoids geometric fractures or topological errors caused by translation, and makes the boundary representation three-dimensional model generated after movement have a complete and continuous topological relationship, thereby improving the quantity and quality after translation processing, and providing a more stable and reliable geometric basis for subsequent feature classification tasks.

[0080] According to one embodiment of the present application, the step of randomly selecting a point on the contact surface of the base model includes: traversing each boundary representation surface in the base model, calculating the area of ​​each surface, and finding the surface with the largest area and using it as the target surface; obtaining the geometric surface of the target surface, and calculating the normal vector and parameter range of the target surface, wherein the parameter range is the size range of the surface projection UV; randomly selecting U and V parameters within the parameter range of the target surface, and determining a random point on the target surface based on the U and V parameters. It should be noted that the above-mentioned random point determination process is called the random planar transform on contact face method.

[0081] According to one embodiment of the present application, the TopExp_Explorer traversal method is used to traverse each boundary representation surface in the base model; the BRep_Tool.Surface method is used to obtain the geometric surface of the target surface; and a BRepGProp_Face surface object is created to calculate the normal vector and parameter range of the target surface.

[0082] It can be seen from the above description that the above embodiments of the present application take into account the geometric properties of the surface (such as normal vectors and parameter ranges). This random point selection method based on geometric properties not only improves the flexibility and adaptability of the translation of complex manufacturing feature models, but also provides a more accurate and reliable geometric basis for subsequent geometric operations, ensuring the rationality and diversity of the complex manufacturing feature models after translation.

[0083] According to one embodiment of the present application, the step of completing the movement processing of the complex manufacturing feature model on the base model also includes: calculating and generating a tangent vector perpendicular to the normal vector based on the normal vector of the target surface; calculating and generating a random direction vector based on the normal vector and the tangent vector; multiplying the random direction vector by a preset random number and using the result as a translation vector for translating the complex manufacturing feature model along the direction of the translation vector, and completing the movement processing of the complex manufacturing feature model when the complex manufacturing feature model intersects with the base model. According to one embodiment of the present application, the BRepAlgoAPI_Section algorithm is used to detect whether the complex manufacturing feature model intersects with the base model.

[0084] As can be seen from the above description, the process of the above embodiment of the present application takes into account the geometric properties of the surface, such as the normal vector, parameter range and randomness, to ensure the rationality and diversity of the complex manufacturing feature model after translation.

[0085] (2) Copy

[0086] According to one embodiment of the present application, the step of replicating the complex manufacturing feature model on the base model includes: replicating the complex manufacturing feature model on the base model, so that the replicating process of the complex manufacturing feature model is completed on the base model. According to another embodiment of the present application, according to another embodiment of the present application, the step of replicating the complex manufacturing feature model on the base model includes: replicating the complex manufacturing feature model on one or more boundary representation surfaces of the base model, so that the replicating process of the complex manufacturing feature model is completed on the base model.

[0087] According to one embodiment of the present application, the steps for completing the copying process of the complex manufacturing feature model on the base model include: using the BRepBuilderAPI_Copy method to copy the complex manufacturing feature model; randomly selecting a point on the contact surface of the base model, wherein the contact surface of the base model is the intersection surface of the base model and the complex manufacturing feature model; using the BRepBuilderAPl_Transform method to translate the copied complex manufacturing feature model along the normal vector direction of the point, and when the copied complex manufacturing feature model intersects with the base model, completing the movement process of the copied complex manufacturing feature model, thereby completing the copying process of the complex manufacturing feature model. The movement operation of the copied complex manufacturing feature model is the same as the scheme for moving the complex manufacturing feature model in the above embodiment, and will not be repeated here.

[0088] As can be seen from the above description, the above embodiments of the present application can generate a rich variety of training samples by replicating complex manufacturing feature models on one or more boundary representation surfaces of the base model. These samples are highly consistent and accurate, which helps to train the image classification system to better understand and process complex geometric relationships and spatial configurations, improve the generalization performance of the image classification system, and reduce the cost and time of data set construction.

[0089] (3) Rotation

[0090] According to one embodiment of the present application, the step of rotating the complex manufacturing feature model on the base model includes: rotating the complex manufacturing feature model on the base model according to a preset rotation angle, so that the rotation processing of the complex manufacturing feature model is completed on the base model. According to another embodiment of the present application, the complex manufacturing feature model is rotated multiple times on the base model according to one or more preset rotation angles, so that the rotation processing of the complex manufacturing feature model on the base model is completed.

[0091] As can be seen from the above description, the above embodiment of the present application rotates the complex manufacturing feature model on the base model according to one or more preset rotation angles, and uses the rotated complex manufacturing feature model as a sample to train the image classification system. This can significantly improve the image classification system's ability to understand and generalize different angles and spatial configurations. This method not only enriches the diversity of training data and enhances the image classification system's ability to handle complex geometric relationships, but also improves its accuracy and robustness in practical applications.

[0092] According to one embodiment of the present application, the steps of completing the rotation processing of the complex manufacturing feature model on the base model include: calculating the normal vector of the surface of the intersection area of ​​the base model and the complex manufacturing feature model; randomly selecting a point on the surface, and constructing a rotation axis based on the normal vector of the point; rotating the complex manufacturing feature model around the rotation axis according to a preset rotation angle to obtain a rotated complex manufacturing feature model. According to one embodiment of the present application, a Boolean intersection operation is performed on the rotated complex manufacturing feature model and the base model to determine whether the rotated complex manufacturing feature model intersects with the base model. If the rotated complex manufacturing feature model does not intersect with the base model, different rotation angle parameters will be tried until the complex manufacturing feature model intersects with the base model or the maximum number of attempts is reached.

[0093] According to one embodiment of the present application, the normal vector of a point is obtained using the Normal method of the BRepGProp_Face object; and the complex manufacturing feature model is rotated around the rotation axis using the gp_Trsf transformation object.

[0094] As can be seen from the above description, the above embodiments of the present application utilize the geometric properties of the surface (such as the normal vector of the surface) to ensure the rationality of the rotation operation, and the establishment of the rotation axis prevents the complex manufacturing feature model from deforming or detaching from the base model. At the same time, the rotated complex manufacturing feature model and the base model are subjected to a Boolean intersection operation to ensure that the complex manufacturing feature model intersects with the base model. This random rotation method based on the geometric properties of the surface normal vector can not only increase the diversity of the complex manufacturing feature model, but also maintain the contact relationship between the complex manufacturing feature model and the base model, thereby ensuring that the geometric topological continuity between the rotated complex manufacturing feature model and the base model is maintained, effectively avoiding geometric breaks or topological errors caused by rotation, and making the boundary representation of the three-dimensional model generated after rotation have a complete and continuous topological relationship, thereby improving the quantity and quality after rotation processing, and providing a more stable and reliable geometric basis for subsequent feature classification tasks.

[0095] (4) Zoom

[0096] According to an embodiment of the present application, the step of scaling the complex manufacturing feature model on the base model includes: scaling the complex manufacturing feature model on the base model to complete the scaling of the complex manufacturing feature model on the base model.

[0097] It can be seen from the above description that the above embodiment of the present application scales the complex manufacturing feature model on the base model. This method not only increases the diversity of the data, but also uses the scaled complex manufacturing feature model as a sample to train the image classification system, which can significantly enhance the image classification system's understanding ability and generalization performance of manufacturing features of different sizes and proportions.

[0098] According to one embodiment of the present application, the step of scaling the complex manufacturing feature model on the base model includes: calculating the centroid of the intersection area of ​​the base model and the complex manufacturing feature model; scaling the complex manufacturing feature model according to a preset scaling factor and the centroid to obtain a scaled complex manufacturing feature model. According to one embodiment of the present application, the gp_Trsf transformation object scales the complex manufacturing feature model; and performing a Boolean intersection operation on the scaled complex manufacturing feature model and the base model to determine whether the scaled complex manufacturing feature model intersects with the base model.

[0099] As can be seen from the above description, the above embodiment of the present application utilizes the geometric properties of the surface (such as the center of mass of the surface) to ensure the rationality of the scaling operation. At the same time, the scaled complex manufacturing feature model is subjected to a Boolean intersection operation with the base model to ensure that the scaled complex manufacturing feature model intersects with the base model. This random scaling method based on the geometric properties of the surface normal vector can not only increase the diversity of the complex manufacturing feature model, but also maintain the contact relationship between the complex manufacturing feature model and the base model, thereby ensuring that the geometric topological continuity between the scaled complex manufacturing feature model and the base model is maintained, effectively avoiding geometric breaks or topological errors caused by scaling, and making the boundary representation of the three-dimensional model generated after scaling have a complete and continuous topological relationship, thereby improving the quantity and quality after rotation processing, and providing a more stable and reliable geometric basis for subsequent feature classification tasks.

[0100] The experimental results of the inventor moving, copying, rotating, and / or scaling multiple complex manufacturing feature models on the base model are shown in the following figure:

[0101] like Figure 2 As shown, it is a schematic diagram of the first example of the base model and the complex manufacturing feature model, wherein: Figure 1 The whole is a supporting bracket, arrow 1 points to the supporting component, which is the base model; arrow 2 points to the screw, arrow 3 points to the hexagonal inner hole cylindrical sleeve, arrow 4 points to the block connector, and arrows 1-3 are all complex manufacturing feature models.

[0102] The three complex manufacturing feature models of screws, hexagonal cylindrical sleeves, and block connectors are translated on the support components to obtain Figure 3 The schematic diagram shown in the figure is as follows; the three complex manufacturing feature models of screws, hexagonal inner hole cylindrical sleeves, and block connectors are copied on the support component to obtain Figure 4 The schematic diagram shown in the figure is as follows: the three complex manufacturing feature models of screws, hexagonal inner hole cylindrical sleeves, and block connectors are rotated on the support component to obtain Figure 5 The schematic diagram shown in the figure is as follows; the three complex manufacturing feature models of screws, hexagonal inner hole cylindrical sleeves, and block connectors are scaled on the support component to obtain Figure 6 The schematic diagram shown.

[0103] like Figure 7 As shown, it is a schematic diagram of the second example of the base model and the complex manufacturing feature model, wherein, Figure 7 The whole is the dashboard of the digital model of the car. The arrow 5 points to the large flat base, which is the base model. The three complex manufacturing feature models such as screws, hexagonal inner hole cylindrical sleeves, and block connectors are translated on the large flat base to obtain Figure 8The schematic diagram shown in the figure is as follows; the three complex manufacturing feature models of screws, hexagonal inner hole cylindrical sleeves, and block connectors are copied on the large flat base to obtain Figure 9 The schematic diagram shown in the figure is as follows: the three complex manufacturing feature models, namely the screw, the hexagonal inner hole cylindrical sleeve and the block connector, are rotated on the large flat base to obtain Figure 10 The schematic diagram shown in the figure is as follows; the three complex manufacturing feature models of screws, hexagonal inner hole cylindrical sleeves, and block connectors are scaled on the large flat base to obtain Figure 11 The schematic diagram shown.

[0104] In addition, according to the three-dimensional model processing method proposed in the above embodiment of the present application, the present application proposes a training method for an image classification system, the image classification system is used to classify the boundary representation surface in the boundary representation three-dimensional model, such as Figure 12 As shown, the method includes: T1, obtaining a boundary representation three-dimensional model; T2, using the three-dimensional model processing method of any of the above embodiments to process the boundary representation three-dimensional model, and forming a data set with the processed results; T3, using the data set to perform multiple iterative training on the image classification system until convergence.

[0105] The inventors conducted a series of experiments on the above embodiments, wherein Figure 13 As shown in the figure, it is the experimental result of data enhancement and data labeling for the dashboard of the digital model of the car. Among them, the component in the red circle is the guide rail slider, the component in the yellow circle is the L-shaped bracket, and the component in the blue circle is the gusset plate with reinforcement ribs. The guide rail slider, L-shaped bracket, and gusset plate with reinforcement ribs are all complex manufacturing feature models; the porous flat plate base below the guide rail slider, L-shaped bracket, and gusset plate with reinforcement ribs is the base model.

[0106] Figure 13 (a) and 13(b) are the results of data enhancement processing on a porous flat base using three complex manufacturing feature models, including a guide rail slider, an L-shaped bracket, and a gusset with reinforcement ribs. Figure 13 (a) is the result after the guide rail slider and L-shaped bracket are copied, and 13(b) is the result after the guide rail slider, L-shaped bracket, and gusset plate with reinforcing ribs are moved.

[0107] Figure 13 (c) is the experimental data related to the annotation of the boundary representation surface. Figure 13(c) Labels for complex manufacturing feature models, such as guide rails, L-shaped brackets, and gussets with ribs. The label format is: boundary surface index: label number. A label number of 0 indicates that the boundary surface belongs to the porous plate base, 1 to the L-shaped bracket, 2 to the guide rail, and 3 to the gusset. In other words, the label indicates which original part the boundary surface belongs to in the target boundary representation 3D model.

[0108] Experimental results show that the image classification system trained by the above embodiments of the present application can accurately identify complex manufacturing features such as gussets with reinforcing ribs, guide rail sliders, L-shaped brackets, clips, screws, boss columns, reinforcing ribs, etc. on complex B-rep three-dimensional models, with an identification accuracy rate of more than 90%.

[0109] From the above description, it can be seen that by using the three-dimensional model processing method proposed in the above embodiment of this application to process the acquired boundary representation three-dimensional model, the number and diversity of manufacturing features in the boundary representation three-dimensional model can be increased. Therefore, when the boundary representation three-dimensional model obtained by the processing scheme of this application is used to train the image classification system, the generalization ability of the image classification system and the classification accuracy of the manufacturing feature model in the boundary representation three-dimensional model can be improved.

[0110] Among them, all the boundary representation three-dimensional models in the inventor's experiments were separated from the front interior and its display on the digital car model (the front interior and the display are both manufacturing feature models of the digital car model). The processing method of the embodiment of the present application on the digital car model can increase the number and diversity of manufacturing feature models such as screws, hexagonal cylindrical sleeves with inner holes, and block connectors in the digital car model, thereby providing enough high-quality training samples for the image classification system that needs to be trained. The three-dimensional model processing method of the present application is not limited to processing only the digital car model. Other mechanical parts that require CAE software for simulation testing to improve their corresponding process products can also adopt the processing method of the present application, which is not specifically limited here.

[0111] According to the training method of the above embodiment of the present application, the present application proposes an application method of an image classification system, such as Figure 14 As shown, the method includes: P1, obtaining a boundary representation three-dimensional model, wherein the boundary representation three-dimensional model includes a boundary representation surface; P2, using the image classification system trained by the training method of the image classification system of the above embodiment to classify the boundary representation surface in the boundary representation three-dimensional model.

[0112] As can be seen from the above description, the above embodiment of the present application utilizes the image classification system trained by the present application to classify the boundary representation surfaces in the boundary representation three-dimensional model, which can improve the classification accuracy of the boundary representation surfaces.

[0113] According to one embodiment of the present application, the image classification system trained in the above embodiment is configured in CAE software, so that the CAE software can perform performance simulation tests on the manufacturing feature model in the boundary representation three-dimensional model.

[0114] From the above description, it can be seen that applying the image classification system to CAE software enables CAE software to reasonably allocate computing resources for its performance simulation test based on the accurately classified manufacturing feature model, thereby enabling CAE software to perform accurate and efficient performance simulation tests on the manufacturing feature model, thereby improving the production efficiency and quality of process products.

[0115] It should be noted that the boundary representation three-dimensional models processed in this application can be used for engineering part models such as automotive engine components and aerospace structural components. Specifically, the above-mentioned embodiments proposed in this application can be used to improve the efficiency and quality of the production of automotive engine components and aerospace structural components. The specific model structure processed is not limited to this. The technical solutions of this application can be used to produce specific mechanical products as needed, and are not specifically limited here.

[0116] In summary, compared to the existing scheme for classifying the manufacturing feature models of the boundary representation three-dimensional model based on the deep learning method, the above embodiment of the present application obtains the target boundary representation three-dimensional model based on the preset complex manufacturing feature model and the boundary representation three-dimensional model. The present application adds a complex manufacturing feature model to the boundary representation three-dimensional model containing the simple structure manufacturing feature model to synthesize the boundary representation three-dimensional model containing the simple structure and complex structure manufacturing feature models. The three-dimensional model processing scheme of the present application adds a complex manufacturing feature model to the boundary representation three-dimensional model. Therefore, when the image classification system is trained using the boundary representation three-dimensional model obtained by the processing scheme of the present application, the generalization ability of the image classification system and the classification accuracy of the manufacturing feature model in the boundary representation three-dimensional model can be improved, thereby improving the production efficiency and quality of the process products.

[0117] Based on the inventive concept of the above embodiments, the present application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the method described in the above embodiments when executing the computer program. Figure 15 Provide detailed explanation.

[0118] like Figure 15 As shown in FIG, the electronic device 100 of the present application is shown, which may specifically include a processor 110 and a memory 120. The memory 120 is coupled to the processor 110.

[0119] The processor 110 is used to control the operation of the electronic device. The processor 110 may also be referred to as a CPU (Central Processing Unit). The processor 110 may be an integrated circuit chip with signal processing capabilities. The processor 110 may also be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic device, a discrete gate or transistor logic device, or a discrete hardware component. A general-purpose processor may be a microprocessor, or the processor 110 may be any conventional processor.

[0120] The memory 120 is used to store computer programs and can be RAM, ROM, or other types of storage terminals. Specifically, the memory 120 may include one or more computer-readable storage media, which may be non-transitory or transient. The memory 120 may also include high-speed random access memory and non-volatile memory, such as one or more disk storage terminals and flash memory storage terminals. In some embodiments, the non-transitory computer-readable storage medium in the memory 120 is used to store at least one program code.

[0121] The processor 110 is configured to execute the computer program stored in the memory 120 to implement the methods described in the various method embodiments of the present application.

[0122] In some embodiments, the electronic device may further include a peripheral terminal interface 130 and at least one peripheral terminal. The processor 110, memory 120, and peripheral terminal interface 130 may be connected via a bus or signal lines. Each peripheral terminal may be connected to the peripheral terminal interface 130 via a bus, signal lines, or circuit boards. Specifically, the peripheral terminal includes at least one of a radio frequency circuit 140, a display screen 150, an audio circuit 160, and a power supply 170.

[0123] The peripheral terminal interface 130 can be used to connect at least one peripheral terminal related to I / O (Input / Output) to the processor 110 and the memory 120. In some embodiments, the processor 110, the memory 120, and the peripheral terminal interface 130 are integrated on the same chip or circuit board; in some other implementations, any one or two of the processor 110, the memory 120, and the peripheral terminal interface 130 can be implemented on separate chips or circuit boards, which is not limited in this embodiment.

[0124] The RF circuit 140 is used to receive and transmit RF (Radio Frequency) signals, also known as electromagnetic signals. The RF circuit 140 communicates with communication networks and other IoT devices via electromagnetic signals, and is therefore the communication circuitry of the electronic device. The RF circuit 140 converts electrical signals into electromagnetic signals for transmission, or converts received electromagnetic signals into electrical signals. Optionally, the RF circuit 140 includes an antenna system, an RF transceiver, one or more amplifiers, a tuner, an oscillator, a digital signal processor, a codec chipset, an operator identification module card, and the like. The RF circuit 140 can communicate with other terminals via at least one wireless communication protocol. Such wireless communication protocols include, but are not limited to, the World Wide Web, metropolitan area networks, intranets, various generations of mobile communication networks (2G, 3G, 4G, and 5G), wireless local area networks, and / or WiFi (Wireless Fidelity) networks. In some embodiments, the RF circuit 140 may also include circuitry related to Near Field Communication (NFC), although this application does not limit this.

[0125] The display screen 150 is used to display a user interface (UI). This UI may include graphics, text, icons, videos, or any combination thereof. When the display screen 150 is a touch screen display, it is also capable of collecting touch signals on or above the surface of the display screen 150. These touch signals can be input as control signals to the processor 110 for processing. In this case, the display screen 150 can also be used to provide virtual buttons and / or a virtual keyboard, also known as soft buttons and / or a soft keyboard. In some embodiments, there may be a single display screen 150, located on the front panel of the electronic device; in other embodiments, there may be at least two display screens 150, located on different surfaces of the electronic device or in a foldable design; in still other embodiments, the display screen 150 may be a flexible display, located on a curved or foldable surface of the electronic device. Furthermore, the display screen 150 may be configured as a non-rectangular, irregular shape, also known as a special-shaped screen. The display screen 150 may be made of materials such as LCD (Liquid Crystal Display) and OLED (Organic Light-Emitting Diode).

[0126] The audio circuit 160 may include a microphone and a speaker. The microphone is used to collect sound waves from the operator and the environment, and convert the sound waves into electrical signals and input them into the processor 110 for processing, or input them into the radio frequency circuit 140 to achieve voice communication. For the purpose of stereo acquisition or noise reduction, there can be multiple microphones, which are respectively set in different parts of the electronic device. The microphone can also be an array microphone or an omnidirectional acquisition microphone. The speaker is used to convert the electrical signal from the processor 110 or the radio frequency circuit 140 into sound waves. The speaker can be a traditional thin film speaker or a piezoelectric ceramic speaker. When the speaker is a piezoelectric ceramic speaker, it can not only convert the electrical signal into sound waves audible to humans, but also convert the electrical signal into sound waves inaudible to humans for purposes such as ranging. In some embodiments, the audio circuit 160 may also include a headphone jack.

[0127] Power supply 170 is used to power various components in the electronic device. Power supply 170 can be AC ​​power, DC power, a disposable battery, or a rechargeable battery. When power supply 170 includes a rechargeable battery, the rechargeable battery can be a wired rechargeable battery or a wireless rechargeable battery. A wired rechargeable battery is a battery that is charged via a wired line, while a wireless rechargeable battery is a battery that is charged via a wireless coil. The rechargeable battery can also be used to support fast charging technology.

[0128] For a detailed description of the functions and execution processes of the various functional modules or components in the electronic device embodiments of the present application, reference can be made to the descriptions in the above-mentioned method embodiments of the present application, which will not be repeated here.

[0129] In the several embodiments provided in this application, it should be understood that the disclosed electronic devices and methods can be implemented in other ways. For example, the various embodiments of the electronic devices described above are merely illustrative. For example, the division of modules or units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some data can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0130] Units described as separate components may or may not be physically separate, and components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0131] In addition, the functional units in the various embodiments of the present application may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0132] Based on the inventive concept of the above embodiments, the present application further provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the steps of the method described in any of the above embodiments are described. Figure 16 The execution process of the above embodiment in a computer-readable storage medium is described.

[0133] like Figure 16 As shown, it shows the computer-readable storage medium of the present application. If the above-mentioned integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium 200. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions / computer programs for making an IoT device (which can be a personal computer, server, or network terminal, etc.) or a processor (processor) perform all or part of the steps of the various embodiments of the present application. The aforementioned storage medium includes various media such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, as well as electronic terminals such as computers, mobile phones, laptops, tablet computers, cameras, etc. having the above-mentioned storage medium.

[0134] The description of the execution process of the program data in the computer-readable storage medium can refer to the description in the above-mentioned method embodiments of the present application, and will not be repeated here.

[0135] The above description is merely an embodiment of the present application and does not limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made using the contents of the present application specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present application.

[0136] Those skilled in the art will understand that in the above-mentioned method of the specific implementation method, the writing order of each step does not mean a strict execution order and does not constitute any limitation on the implementation process. The specific execution order of each step should be determined by its function and possible internal logic.

Claims

1. A method for processing a three-dimensional model, the method being used for processing a boundary representation three-dimensional model, characterized in that: The method comprises: Obtaining a boundary representation three-dimensional model; Based on the boundary representation three-dimensional model and the preset complex manufacturing feature model, a target boundary representation three-dimensional model is calculated.

2. The three-dimensional model processing method according to claim 1, characterized in that: The boundary representation three-dimensional model includes: a base model and a manufacturing feature model, and the step of calculating and obtaining a target boundary representation three-dimensional model based on the boundary representation three-dimensional model and the preset complex manufacturing feature model includes: The complex manufacturing feature model is moved, and / or copied, and / or rotated, and / or scaled on the base model, and the processed complex manufacturing feature model and the manufacturing feature model are fused with the base model to obtain the target boundary representation three-dimensional model.

3. The three-dimensional model processing method according to claim 2, characterized in that: The step of moving the complex manufacturing feature model on the base model comprises: The complex manufacturing feature model is translated on the base model to complete the movement process of the complex manufacturing feature model on the base model.

4. The three-dimensional model processing method according to claim 2, characterized in that: The step of replicating the complex manufacturing feature model on the base model comprises: The complex manufacturing feature model is copied on the base model to complete the copying process of the complex manufacturing feature model on the base model.

5. The three-dimensional model processing method according to claim 2, characterized in that: The step of rotating the complex manufacturing feature model on the base model comprises: The complex manufacturing feature model is rotated on the base model according to a preset rotation angle, so that the rotation processing of the complex manufacturing feature model is completed on the base model.

6. The three-dimensional model processing method according to claim 2, characterized in that: The step of scaling the complex manufacturing feature model on the base model comprises: The complex manufacturing feature model is scaled on the base model, so that the scaling process of the complex manufacturing feature model is completed on the base model.

7. A training method for an image classification system for classifying boundary representation surfaces in a boundary representation three-dimensional model, characterized in that: The method comprises: Obtaining a boundary representation three-dimensional model; Using the three-dimensional model processing method according to claim 1, the boundary representation three-dimensional model is processed, and the processed results form a data set; The image classification system is trained multiple times using the dataset until convergence.

8. An application method of an image classification system, characterized in that: The method comprises: Acquire a boundary representation three-dimensional model, wherein the boundary representation three-dimensional model includes a boundary representation surface; The image classification system trained by the training method of the image classification system according to claim 7 is used to classify the boundary representation surfaces in the boundary representation three-dimensional model.

9. An electronic device, characterized in that: The method comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the method according to any one of claims 1 to 6, 7, and 8 when executing the computer program.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6, 7 and 8 are implemented.

Images