Non-structural functional feature identification method, device and storage medium

Abstract

The application discloses a non-structural functional feature identification method, device and storage medium, and relates to the technical field of computer-aided engineering preprocessing. The non-structural functional feature identification method comprises the following steps: determining a reference surface currently traversed, and judging whether the reference surface has an inner ring based on the topological property of the reference surface; if the reference surface has an inner ring, searching for a face sheet set connected with the inner ring to determine a connection feature; and determining a design element corresponding to the reference surface based on the concave-convex property of the connection feature. The application solves the technical problem that related technologies are relatively sensitive to geometric shapes and sizes, leading to poor robustness of feature identification, and achieves the technical effect of cross-face sheet non-structural functional feature identification.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer aided engineering preprocessing, and particularly relates to a non-structural functional feature identification method, device and storage medium. BACKGROUND

[0002] Non-structural functional features, such as signs, decorative reliefs, process holes, grooves, ribs, bosses and the like, are widely present in complex geometrical bodies. These design element type features have significance in visual presentation or process manufacturing in the product design stage, but in structural simulation analysis, they will destroy the topological continuity of the geometric model, causing problems such as middle surface extraction failure, mesh quality degradation, and cell number explosion, thereby greatly affecting the simulation efficiency and result accuracy.

[0003] The feature identification function in the related art relies on heuristic rules or geometric template matching, and is relatively sensitive to geometric shapes and sizes, resulting in significant limitations of the identification result, poor robustness under complex geometric conditions, and inaccurate feature identification result. SUMMARY

[0004] The main purpose of the present application is to provide a non-structural functional feature identification method, device and storage medium, aiming to solve the technical problem that the related art is relatively sensitive to geometric shapes and sizes, resulting in poor robustness of feature identification.

[0005] To achieve the above purpose, the present application provides a non-structural functional feature identification method, which comprises:

[0006] determining a reference surface currently traversed, and judging whether the reference surface has an inner ring based on the topological nature of the reference surface;

[0007] if the reference surface has an inner ring, searching for a feature surface associated with the inner ring to determine a connection feature;

[0008] determining a design element corresponding to the reference surface based on the concave-convex nature of the connection feature.

[0009] In an embodiment, the determination of the reference surface currently traversed comprises:

[0010] determining the reference surface currently traversed according to a face patch identifier; or

[0011] determining an initial reference surface currently traversed according to a face patch identifier;

[0012] Determining, in a neighborhood of the initial reference plane, a neighboring facet that satisfies a smooth connectivity requirement with the initial reference plane, wherein a first normal angle between the facet in the neighborhood and the initial reference plane is less than a preset angle, and / or a change in boundary curvature between the facet in the neighborhood and the initial reference plane is less than a preset threshold, and determining that the facet is a neighboring facet that satisfies the smooth connectivity requirement;

[0013] The reference plane is obtained by removing shared internal edges in an extended base plane region formed by the initial reference plane and the neighboring face patches.

[0014] In one embodiment, removing the shared internal edges in the extended base surface region formed by the initial reference surface and the neighboring facets to obtain the reference surface includes:

[0015] Determining edges in an extended base surface area formed by the initial reference surface and the neighboring surface patches;

[0016] Determine isolated edges in the edge lines, and determine the edges other than the isolated edges in the edge lines as shared internal edges; or determine the shared internal edges in the edge lines, and determine the edges other than the shared internal edges in the edge lines as isolated edges;

[0017] The shared internal edges in the extended base surface area are removed to obtain the reference surface.

[0018] In one embodiment, determining the isolated edge in the edge line includes:

[0019] constructing an undirected edge structure based on the hash values ​​of the edges;

[0020] Counting the number of occurrences of each edge in the undirected edge structure according to a hash table;

[0021] The target edge line whose number of occurrences is a preset value is determined to be the isolated edge.

[0022] In one embodiment, determining whether the reference plane has an inner ring based on the topology of the reference plane includes:

[0023] Performing closed loop search based on isolated edges in the reference plane to determine a closed boundary loop;

[0024] If the number of the closed boundary rings is greater than two, it is determined that an inner ring exists on the reference plane.

[0025] In one embodiment, searching for the characteristic surface associated with the inner ring to determine the connection feature includes:

[0026] Extracting target facets directly adjacent to the boundary according to the topological connection relationship of the boundary of the inner ring;

[0027] constructing an initial feature face set according to the target face;

[0028] iteratively searching extended face based on geometric continuity and topological connectivity of the initial feature face set, and taking the initial feature face and the extended face as the connecting feature, wherein the initial feature face and the extended face are the feature faces.

[0029] In an embodiment, the connecting feature includes a reference face and a feature face, and before determining the design element corresponding to the reference face based on the concave-convex of the connecting feature, the method further includes:

[0030] determining a second normal angle between the average normal of the reference face in the connecting feature and the central normal of each feature face,

[0031] determining an average angle of the connecting feature according to the second normal angle;

[0032] determining the concave-convex of the connecting feature based on the average angle.

[0033] In an embodiment, the determining the design element corresponding to the reference face based on the concave-convex of the connecting feature includes:

[0034] determining a geometric profile of the connecting feature based on geometric parameters of the connecting feature;

[0035] identifying the design element in the model according to the concave-convex and the geometric profile.

[0036] In addition, to achieve the above object, the present application also provides a non-structural functional feature recognition device, which comprises a memory, a processor and a computer program stored in the memory and executable on the processor, and the computer program is configured to implement the steps of the non-structural functional feature recognition method.

[0037] In addition, to achieve the above object, the present application also provides a storage medium, which is a computer readable storage medium, and a program for implementing a non-structural functional feature recognition method is stored on the computer readable storage medium, and the program for implementing a non-structural functional feature recognition method is executed by a processor to implement the steps of the non-structural functional feature recognition method.

[0038] The application provides a non-structural functional feature recognition method, and first determines a reference surface currently traversed, and judges whether the reference surface has an inner ring based on the topological property of the reference surface; if the reference surface has an inner ring, a connected feature is determined by searching a set of surface patches connected with the inner ring; and a design element corresponding to the reference surface is determined based on the concave-convex property of the connected feature. That is, the application constructs a set of surface patches based on the reference surface where the inner ring is located according to the topological property of the inner ring, and then extracts a connected feature, and then recognizes a design element according to the concave-convex property of the connected feature. The technical problem of insufficient robustness of feature recognition caused by strong sensitivity of related technologies to geometric shapes and sizes is solved, and the technical effect of cross-surface non-structural functional feature recognition is achieved. BRIEF DESCRIPTION OF DRAWINGS

[0039] The accompanying drawings, which are incorporated into and form a part of the specification, illustrate an embodiment consistent with the application and, together with the description, serve to explain the principles of the application.

[0040] In order to more clearly illustrate the technical solutions in the embodiments of the application or the prior art, brief introductions will be given to the drawings needed to be used in the embodiments or prior art descriptions. Obviously, for those skilled in the art, other drawings can also be obtained from these drawings without creative labor.

[0041] Figure 1 A flowchart is provided for the non-structural functional feature recognition method embodiment one of the application;

[0042] Figure 2 An example diagram is provided for deleting shared internal edges in the non-structural functional feature recognition method embodiment two of the application;

[0043] Figure 3 An example of non-structural functional feature recognition in the non-structural functional feature recognition method embodiment six of the application is provided;

[0044] Figure 4 Another example of non-structural functional feature recognition in the non-structural functional feature recognition method embodiment six of the application is provided;

[0045] Figure 5 Still another example of non-structural functional feature recognition in the non-structural functional feature recognition method embodiment six of the application is provided;

[0046] Figure 6 A whole flowchart of the non-structural functional feature recognition method in the embodiment of the application is provided;

[0047] Figure 7 A system structure diagram of the non-structural functional feature recognition device of the application is provided.

[0048] The object, features and advantages of the present application will be further illustrated in conjunction with the embodiments, with reference to the accompanying drawings. DETAILED DESCRIPTION

[0049] It should be understood that the specific embodiments described herein are merely intended to explain the technical solutions of the present application, and are not intended to limit the present application.

[0050] In order to better understand the technical solutions of the present application, the specific embodiments will be described in detail below in conjunction with the accompanying drawings and specific embodiments.

[0051] At present, in the engineering design and simulation scene of CAD (Computer-Aided Design) / CAE (Computer-Aided Engineering), a three-dimensional geometric model usually contains a large number of geometric features, which include not only structural functional features (such as holes, grooves, ribs, bosses, etc.), but also decorative or identifying features (such as text logos, patterns, corporate logos, etc.). Among them, although the logo features do not directly participate in the structural stress, they often appear in the prominent position of the product, such as the surface of the shell, bracket, shell, etc. parts, which has important significance for the product appearance and brand identification.

[0052] However, before performing CAE simulation such as FEA (Finite Element Analysis) or CFD (Computational Fluid Dynamics), the three-dimensional geometric model must be pre-processed. Pre-processing usually includes steps such as geometry cleaning, mid-surface extraction, model simplification, and meshing. Logo features in CAE analysis can cause mid-surface extraction abnormalities, local mesh quality degradation, and simulation convergence difficulties. Therefore, identifying and processing these non-structural geometric features in the pre-processing stage is one of the key steps to ensure simulation efficiency and accuracy.

[0053] The current mainstream CAD / CAE pre-processing software has certain ability in geometric feature recognition, but it has limited recognition ability for complex, non-structural, especially logo structures that span multiple face patches, are nested or embossed. These logos are often embedded in the reference surface through smooth connection of multiple faces, round corner transition, and curvature continuity. Traditional algorithms based on topological structure or single face feature recognition are difficult to adapt to this type of complex structure. In addition, current recognition methods often require users to manually select seed faces, input boundary contours, or use interactive tools to assist in recognition, which cannot be automatically processed in batches, lack the ability to uniformly recognize, filter the size and judge the type of these features, and seriously affect the pre-processing efficiency.

[0054] The main solution of the present application is to determine the reference surface currently traversed, and determine whether the reference surface has an inner ring based on the topological property of the reference surface; if the reference surface has an inner ring, search for the feature surface associated with the inner ring to determine the connecting feature; and determine the design element corresponding to the reference surface based on the concave-convex property of the connecting feature.

[0055] The present application constructs a patch set according to the topological property of the inner ring based on the reference surface where the inner ring is located, further extracts the connecting feature, and then identifies the design element according to the concave-convex property of the connecting feature, thereby achieving the technical effect of cross-patch non-structural functional feature identification.

[0056] It should be noted that the execution subject of the present embodiment can be a terminal device, or a computing service device with data processing, network communication and program running functions, such as a tablet computer, a personal computer, a mobile phone, or a non-structural functional feature identification device capable of achieving the above functions, and the present embodiment does not make specific limitations thereon. The following will take a non-structural functional feature identification device as an example to illustrate the present embodiment and the following embodiments.

[0057] Based on this, the first embodiment of the present application proposes a non-structural functional feature identification method, please refer to Figure 1 , which comprises steps S10-S30:

[0058] Step S10, determine the reference surface currently traversed, and determine whether the reference surface has an inner ring based on the topological property of the reference surface.

[0059] In the present embodiment, the reference surface refers to a surface selected in a three-dimensional model as the basis for identifying design elements, which can be a plane or a weak curved surface. Topological property refers to the connection relationship and structural characteristics between geometric elements such as faces, edges and points in a geometric model. The inner ring refers to a closed boundary surrounded by the face on the reference surface, such as a closed boundary formed after local features such as holes and logos are processed on a plane.

[0060] As an optional implementation, after importing the three-dimensional model, each face of the model is sequentially traversed, the face currently traversed is directly taken as the reference surface, and it is determined whether the reference surface has an inner ring by checking whether the face contains a closed boundary surrounded by itself.

[0061] As another optional implementation, when a face is traversed, based on smooth connectivity of the face, such as normal angle threshold and boundary curvature continuity, neighborhood expansion is performed from the face as a starting point to obtain a plurality of face sheet sets continuously connected therewith, common edges shared between faces in the set are removed, only isolated edges are retained, the face sheet set after the above processing is taken as a reference face, and each face in the set is marked with a read mark, and the marked face is skipped in subsequent traversal, and then whether the reference face has an inner ring is determined by analyzing whether the isolated edges in the reference face can form a closed boundary.

[0062] In step S20, if the reference face has an inner ring, a feature face associated with the inner ring is searched to determine a connection feature.

[0063] In the embodiment, the feature face refers to a face directly connected with the inner ring and jointly constituting a design feature. The connection feature is a structure with a specific geometric shape formed by a plurality of feature faces, such as a hole wall face set constituting a hole feature.

[0064] As an optional implementation, the boundary edges of the inner ring are taken as a starting point, and the topological connection relationship of geometric elements is used to search for face sheets directly connected with the boundary edges, and the searched face sheets are taken as feature faces to form an initial connection feature.

[0065] As another optional implementation, a breadth-first search algorithm is adopted, the inner ring is taken as a center, and geometric continuity, such as face normal angle, common edge relationship, and smooth connectivity, is used to search for faces continuously connected with the inner ring layer by layer outward, and all searched face sets are determined as connection features.

[0066] In step S30, a design element corresponding to the reference face is determined based on the concave-convex property of the connection feature.

[0067] In the embodiment, the concave-convex property is used to describe the convex or concave state of the connection feature relative to the reference face. The design element refers to a geometric structure with a specific function or shape in the three-dimensional model, such as a boss, a hole, a groove, a logo, and the like. The design element is a non-structural functional feature in the embodiment. In contrast, the structural functional feature is a geometric structure in the three-dimensional model that plays a key supporting role in realizing the core mechanical properties and functions of the product, such as structural strength, mechanical properties, assembly relationship, and the like. Such features directly participate in force transmission, load bearing, positioning connection, or spatial constraint, and are basic elements to ensure stable operation of the product under actual working conditions, such as bearing, movement, and environmental adaptation, to meet the mechanical and functional requirements.

[0068] As an optional implementation, the angle between the normal of the connecting feature surface and the normal of the reference surface is calculated. If the average angle is greater than 90 degrees, the connecting feature is determined to be concave, and if the average angle is less than 90 degrees, the connecting feature is determined to be convex. For the convex connecting feature, further identification is performed according to the geometric shape, such as identifying a cylindrical feature as a boss and identifying an elongated cylindrical feature as a rib plate. For the concave connecting feature, it is determined whether the connecting feature penetrates the reference surface. If the connecting feature penetrates the reference surface, the connecting feature is a through hole. If the connecting feature does not penetrate the reference surface, the connecting feature is identified as a blind hole or a groove according to the contour shape.

[0069] As another optional implementation, the local curvature of the boundary between the connecting feature and the reference surface is analyzed. According to the positive and negative of the curvature, the concave-convex property is determined. According to the size of the connecting feature, the contour complexity and other factors, the design element type is further refined, such as a concave feature with a complex contour and not penetrating the reference surface being identified as a logo.

[0070] For example, after the three-dimensional model of the automobile engine cylinder block is imported, each face is traversed in sequence. When a plane on the cylinder block is traversed, it is determined that the plane has an inner ring of a circular shape. By searching the cylindrical surface feature surface associated with the inner ring, it is determined that the connecting feature is a cylindrical structure. By calculating the angle between the normal of the cylindrical surface and the normal of the plane, it is determined that the connecting feature is concave, and the cylindrical structure penetrates the plane. Finally, it is determined that the design element corresponding to the reference surface is a through hole.

[0071] Optionally, after the design element corresponding to the reference surface is determined based on the concave-convex property of the connecting feature, if the currently traversed reference surface is not the last surface, step S10 is continuously executed. If the currently traversed reference surface is the last surface, all the design elements identified in the traversal result are obtained and highlighted on the three-dimensional model.

[0072] In this embodiment, the last surface refers to the last surface in the traversal order among all the surfaces in the three-dimensional model. The traversal result refers to the design element information identified for each reference surface during the execution of steps S10 to S30. Highlighting refers to changing the visual properties of the design element in the three-dimensional model, such as color, contour line thickness, etc., so that it is more eye-catching in the model, facilitating user observation and confirmation.

[0073] As an optional implementation, a surface traversal counter is set in the software system. The counter is incremented by 1 after the processing of each reference surface is completed. When the counter value is equal to the total number of surfaces in the three-dimensional model, the current reference surface is determined to be the last surface. At this time, all the identified design elements are summarized, and the color of the design elements in the three-dimensional model display interface is modified by the graphics rendering engine, for example, the boss feature is displayed in red and the hole feature is displayed in blue.

[0074] As another optional implementation, in the traversal process, each recognized design element information is stored in a data list, when the last face is traversed, the data list is read, and according to the design element coordinates, geometric shapes and other information in the list, a highlighted contour line is added for each design element in the visualization interface of the three-dimensional model, and a design element list is generated on the interface for the user to quickly view and retrieve.

[0075] For example, after the three-dimensional model of the automobile engine cylinder body is imported, each face is traversed in turn. When a plane on the cylinder body is traversed, it is judged that the plane has an inner ring of a circular shape, the connecting feature of the inner ring is determined to be a cylindrical structure by searching the cylindrical feature surface associated with the inner ring, and the connecting feature is judged to be concave by calculating the angle between the normal of the cylindrical surface and the normal of the plane, and the cylindrical structure penetrates the plane, and finally the design element corresponding to the reference plane is determined to be a through hole. Other faces are continuously traversed, and design elements such as bosses and rib plates on the cylinder body are identified. When all the faces are traversed, the identified through holes, bosses, rib plates and other design elements are highlighted in different colors on the three-dimensional model, the through holes are displayed in blue, the bosses are displayed in red, and the rib plates are displayed in green, which facilitates engineers to intuitively view the distribution of design elements in the model.

[0076] The embodiment is based on the reference plane where the inner ring is located, constructs a patch set according to the topological property of the inner ring, and then extracts the connecting feature, and finally identifies the design element according to the concave-convex property of the connecting feature. The technical problem of insufficient robustness of feature recognition caused by strong sensitivity to geometric shapes and sizes in related technologies is solved, and the technical effect of cross-patch non-structural functional feature recognition is achieved.

[0077] Based on any of the above embodiments, in the second embodiment of the present application, the reference plane currently traversed is determined, including: step S11, or steps S12-S14.

[0078] Step S11, determining the reference plane currently traversed according to the patch identifier.

[0079] In this embodiment, the patch identifier is a number or label assigned to each patch based on the model data structure.

[0080] As an optional implementation, after the three-dimensional model is imported, the patch identifiers are read in a predetermined order (such as from the starting position of the model data storage), and the patch corresponding to the currently read identifier is directly determined as the reference plane. If the patch is read for the first time, it is the first patch of the entire model; if it is not the first time, it is the current untraversed patch in the traversal order.

[0081] As another optional implementation, a patch identification index table is constructed, and the patch corresponding to the patch identification pointed by the pointer is determined as the reference patch in sequence by moving the pointer in the index table. When the pointer is at the start position of the index table, the first patch of the model is determined; when the pointer is at the middle position, the current untraversed patch in the traversal process is determined.

[0082] In step S12, the initial reference patch currently traversed is determined according to the patch identification.

[0083] In the embodiment, the initial reference patch is the start patch in the reference patch determination process.

[0084] As an optional implementation, after the model is imported, the first patch is selected as the initial reference patch from the head of the model patch set according to the patch identification. If it is the first traversal, the initial reference patch is the first patch of the model; if it is subsequent traversal, the initial reference patch is selected from the head of the untraversed patch set.

[0085] As another optional implementation, a specific screening condition is set, such as the largest area or the coordinates in a specific range, and the patch meeting the condition is screened from the untraversed patch set, and the patch corresponding to the patch identification of the patch is determined as the initial reference patch.

[0086] As another optional implementation, each patch is marked with a patch identification, and the patch identification has a numerical size. The patches are traversed in a preset order according to the numerical size of the patch identification, and the preset order can be from small to large or from large to small.

[0087] Further, if the determined reference patch contains multiple patches, the patch identifications of the contained patches are all marked as read, and then when the design element recognition of the reference patch is completed, the next initial reference patch is traversed, and the read patches are skipped.

[0088] In step S13, the neighborhood patch in the neighborhood of the initial reference patch that meets the smooth connection requirement with the initial reference patch is determined.

[0089] In the embodiment, the neighborhood patch refers to the patch set adjacent to the initial reference patch, and the smooth connection requirement is judged based on the double conditions of the normal angle and the boundary curvature change. If the first normal angle between the patch in the neighborhood and the initial reference patch is less than a preset angle, and / or the boundary curvature change between the patch in the neighborhood and the initial reference patch is less than a preset threshold, the patch is determined as the neighborhood patch meeting the smooth connection requirement.

[0090] As an optional implementation, the first normal angle between each patch in the neighborhood and the initial reference patch is calculated, and when the angle is less than a preset angle, such as 10 degrees, and the boundary curvature change is less than a preset threshold, the patch is determined as the neighborhood patch meeting the requirement.

[0091] As another optional implementation, the boundary curvature variation of the neighborhood patch and the initial reference surface is determined first. If the variation is less than a preset threshold, then it is determined whether the normal angle is less than a preset angle. If both conditions are met, the neighborhood patch is determined to meet the smooth connection requirement.

[0092] As another optional implementation, the first normal angle between each patch in the neighborhood and the initial reference surface is calculated. If the angle is less than a preset angle, such as 10 degrees, the patch is determined to meet the requirement.

[0093] As another optional implementation, the boundary curvature variation of the neighborhood patch and the initial reference surface is determined. If the variation is less than a preset threshold, the patch is determined to meet the requirement.

[0094] In step S14, the shared internal edges in the extended surface region composed of the initial reference surface and the neighborhood patches are removed to obtain the reference surface.

[0095] In this embodiment, the extended surface region is a set of surfaces composed of the initial reference surface and the neighborhood patches that meet the conditions; the shared internal edge is a repeated connecting edge between adjacent patches in the region.

[0096] As an optional implementation, all edges of the extended surface region are stored as undirected edge structures, the number of occurrences of each edge is counted using a hash table, the shared internal edges with a number of occurrences greater than 1 are removed, and the set of patches corresponding to the remaining edges is the reference surface.

[0097] As another optional implementation, the edges of the extended surface region are traversed. If an edge belongs to two adjacent patches, it is marked as a shared internal edge and deleted. Finally, the reference surface is obtained.

[0098] For example, in the processing of a three-dimensional model of an aero-engine blade, a plane on the leading edge of the blade is selected as the initial reference surface based on the patch identifier during the first traversal; and a curved surface is selected as the initial reference surface from the set of untraversed patches in the trailing edge region during subsequent traversal. For the initial reference surface in the trailing edge region, it is determined that there are three curved surfaces in the neighborhood that meet the normal angle and boundary curvature conditions, and they are included in the neighborhood patches. After removing the shared internal edges, the reference surface containing the complete transition structure is determined, providing an accurate basis for subsequent feature recognition.

[0099] This embodiment forms a hierarchical and progressive reference surface construction process by refining the reference surface determination step. It not only retains the efficiency of directly determining the reference surface, but also enhances the processing capability for complex curved surfaces and smooth transition structures through the expansion and screening mechanism, avoiding feature recognition omission or misjudgment caused by improper selection of the reference surface, and providing a more accurate and reliable geometric basis for subsequent non-structural functional feature recognition.

[0100] Optionally, step S14 comprises:

[0101] Step S141, determining the edge line in the extended base surface region formed by the initial base surface and the adjacent patches.

[0102] In this embodiment, the edge line refers to the set of boundary edges of all patches in the extended base surface region.

[0103] As an optional implementation, the terminal device loads the three-dimensional model data into the memory and constructs a double-linked list structure containing patch IDs and boundary edges. The processor iterates through the initial base surface and its adjacent patches, extracts the boundary edges of each patch through the linked list pointer, stores them in the dynamic array edgeList, and generates a unique identifier EdgeID for each edge and records the patch number to which it belongs.

[0104] As another optional implementation, the terminal device uses a hash table faceEdgeMap to store the mapping relationship between patches and boundary edges, with the patch ID as the key and the corresponding boundary edge set as the value. The processor iterates through the patches in the extended base surface region, merges the boundary edges into the global edge set globalEdgeSet, and generates the edge line list after de-duplication through bit operations.

[0105] Step S142, determining the isolated edge in the edge line to determine the shared internal edge in the edge line other than the isolated edge; or determining the shared internal edge in the edge line to determine the isolated edge in the edge line other than the shared internal edge.

[0106] In this embodiment, the isolated edge refers to an edge that appears only once in the extended base surface region, and the shared internal edge refers to an edge that appears more than once.

[0107] As an optional implementation, the processor of the terminal device creates a hash table edgeCountMap in the memory, with the EdgeID as the key and the edge occurrence count as the value. When iterating through edgeList, if the EdgeID of each edge exists, the count is incremented by 1, otherwise the count is initialized to 1. After the iteration is complete, the edges with a count of 1 are marked as isolated edges and stored in isolatedEdgeList, and the edges with a count greater than 1 are marked as shared internal edges and stored in sharedEdgeList.

[0108] As another optional implementation, the terminal device constructs an edge-patch adjacency matrix edgeFaceMatrix, with the matrix rows and columns representing edges and patches, respectively. If an edge belongs to a patch, the corresponding position is set to 1. The processor calculates the matrix column sum, and edges with a column sum of 1 are determined as isolated edges, and edges with a column sum of 2 are determined as shared internal edges, and are stored in different data structures respectively.

[0109] Step S143, remove the shared internal edge in the extended base surface region, to obtain the reference surface.

[0110] In this embodiment, after removing the shared internal edge, the remaining isolated edge constitutes the boundary of the reference surface.

[0111] As an optional implementation, the processor of the terminal device traverses the sharedEdgeList, and deletes the shared internal edge from the three-dimensional model topology according to the EdgeID. The specific operation includes: updating the face boundary chain table related to the edge, removing the edge reference; if the face boundary is discontinuous due to the deletion of the edge, the boundary ring is reconstructed. After the deletion is completed, the face set corresponding to the remaining isolated edge is encapsulated as the reference surface data object finalBaseSurface.

[0112] As another optional implementation, the terminal device adopts parallel computing optimization processing, and divides the sharedEdgeList into subarrays and distributes them to multi-core processor threads. Each thread independently performs edge deletion operation, and synchronizes the results through semaphore, and finally generates a continuous reference surface topology.

[0113] For example, referring to Figure 2 , Figure 2 An example of deleting a shared internal edge is shown. The left side is the extended base surface region before removing the shared internal edge, and the extended base surface region includes geometric face 1, geometric face 2, geometric face 3, and geometric face 4. For example, the common edge between geometric face 1 and geometric face 2 belongs to both geometric face 1 and geometric face 2. Figure 2 The right side in the middle is the reference surface obtained after removing the shared internal edge. The dashed line is the shared internal edge. After processing the common edge, i.e. the shared internal edge, the remaining edges are combined to form an outer ring and an inner ring. The outer ring is ring 1, and the inner ring is ring 2.

[0114] For example, in the processing of the three-dimensional model of the aero-engine combustion chamber, the terminal device loads the model, and extracts 1258 edge lines of the extended base surface region through the double-linked list. The number of edge occurrences is counted using a hash table, and 189 isolated edges and 1069 shared internal edges are identified. The combustion chamber wall surface is combined into a continuous reference surface.

[0115] This embodiment is based on the hardware architecture and data processing logic of the terminal device. Through the construction of data structures such as double-linked list and hash table, combined with traversal statistics and parallel computing algorithm, the automatic identification and efficient removal of shared internal edges are realized. This scheme reduces manual intervention by using machine computing power, improves the efficiency of reference surface extraction, lays a topological continuity foundation for the identification of non-structural functional features of complex three-dimensional models, and effectively avoids feature recognition errors caused by redundant edges.

[0116] Optionally, determining the isolated edges in the edge lines comprises:

[0117] In step S1421, a hash value of the edge line is constructed.

[0118] In this embodiment, the undirected edge structure is a topological representation without considering the direction of the edge, and the uniqueness of the edge is ensured by the hash value. A unique hash value is generated for each edge. If the hash value of a certain edge appears multiple times, the edge is not isolated. If it appears only once, the edge is isolated.

[0119] As an optional implementation, the processor of the terminal device generates a hash value based on the coordinates of the vertices for each edge line, and the calculation formula is: hash=(vertex1.x+vertex2.x)*(vertex1.y+vertex2.y)*(vertex1.z+vertex2.z), where vertex1 and vertex2 are the two endpoints of the edge. The edges with the same hash value are regarded as the same edge, and an undirected edge object is constructed, which includes the identifier of the edge, the coordinates of the endpoints, and the list of associated facet IDs.

[0120] As another optional implementation, a space hash algorithm is used to map the three-dimensional coordinates of the edge to a two-dimensional grid, and each grid cell stores the hash value of the edge falling into the region. By traversing the grid cells to merge the repeated edges, an adjacency list structure of the undirected edge is constructed.

[0121] In step S1422, the number of occurrences of each edge line in the undirected edge structure is counted according to the hash table.

[0122] In this embodiment, the number of occurrences reflects the situation that the edge is shared by different facets.

[0123] As an optional implementation, the processor of the terminal device creates a hash table edgeCountMap in the memory, with the hash value of the undirected edge as the key and the count of the edge as the value. The undirected edge structure is traversed, and for each edge, if the hash value already exists, the corresponding count is incremented by 1, otherwise the count is initialized to 1.

[0124] In step S1423, the target edge line with the number of occurrences being a preset value is determined as the isolated edge.

[0125] In this embodiment, the preset value is 1, indicating an edge used by only one facet.

[0126] As an optional implementation, the processor traverses edgeCountMap, marks the edges with a count of 1 as isolated edges, and stores them in the isolatedEdgeList array; marks the edges with a count greater than 1 as shared internal edges, and stores them in the sharedEdgeList array.

[0127] For example, in the processing of a three-dimensional model of a mechanical part, for the edge set of the extended base surface area, a hash value of vertex coordinates is used to construct an undirected edge structure, and statistics show that among the 157 edges, 43 edges have an occurrence frequency of 1, and these edges are determined as isolated edges, and the remaining 114 edges are shared internal edges.

[0128] In this embodiment, the efficient identification of isolated edges is achieved by constructing an undirected edge structure and a hash table. This method uses the fast search feature of the hash algorithm to avoid the inefficiency of traditional traversal comparison, and provides an accurate data basis for the subsequent removal of shared internal edges, thereby improving the quality and efficiency of the base surface extraction.

[0129] Based on any of the above embodiments, in the third embodiment of the present application, whether the base surface has an inner ring is determined based on the topological properties of the base surface, including:

[0130] Step A11, based on the isolated edges in the base surface, a closed boundary ring is searched and determined.

[0131] In this embodiment, the base surface refers to a surface selected in the three-dimensional model as the basis for identifying design elements, which can be a plane or a weak curved surface. Topology refers to the connection relationship and structural characteristics between geometric elements such as faces, edges, and points in a geometric model. The inner ring refers to a closed boundary enclosed by the base surface, such as a closed boundary formed by machining a hole, logo, or other local features on a plane. The closed boundary ring is a closed path composed of a series of isolated edges connected end to end. The isolated edge refers to a boundary edge of a base surface that belongs to only one face.

[0132] As an optional implementation, an isolated edge is selected from the set of isolated edges of the base surface as a starting edge, and one end point of the edge is taken as the starting point. The next isolated edge connected to the end point is searched, and the process is repeated until a closed ring is formed. In the search process, the accessed edges and vertices are recorded to avoid repeated access.

[0133] As another optional implementation, an adjacency list of isolated edges is constructed, in which each vertex corresponds to a list storing all isolated edges connected to the vertex. Starting from any vertex, a depth-first search is performed along the edges in the adjacency list until the starting vertex is returned, forming a closed boundary ring.

[0134] As still another optional implementation, a random edge is selected as a seed edge, and the remaining edges are traversed to check whether their end points coincide with the end points of the seed edge. If they coincide, it means that the edges are connected. The loop is traversed until a complete ring, i.e., a closed boundary ring, is found.

[0135] Step A12, if the number of closed boundary rings is greater than two, it is determined that the base surface has an inner ring.

[0136] In this embodiment, the number of closed boundary loops refers to the number of mutually disjoint closed boundary loops determined by the closed loop search.

[0137] As an optional implementation, the number of all found closed boundary loops is counted, and if the number is greater than two, it is considered that the reference surface has an inner ring.

[0138] As another optional implementation, topological analysis is performed on each closed boundary loop to determine its nesting relationship on the reference surface. If one or more closed boundary loops are completely surrounded by other closed boundary loops, it is determined that the reference surface has an inner ring.

[0139] For example, in a flat model with multiple circular holes, three closed boundary loops are determined by closed loop search on the isolated edges of the reference surface. One of the larger closed boundary loops surrounds two smaller closed boundary loops, so it is determined that the reference surface has an inner ring.

[0140] This embodiment can effectively identify the closed boundary loops in the reference surface by closed loop search based on the isolated edges in the reference surface, and accurately determine whether the reference surface has an inner ring by judging the number or nesting relationship of the closed boundary loops. This method fully utilizes the topological characteristics of the reference surface, avoids complex geometric analysis of the entire three-dimensional model, and improves the efficiency and accuracy of inner ring identification.

[0141] Based on any of the above embodiments, in the fourth embodiment of the present application, searching for the feature surface associated with the inner ring to determine the connection feature comprises:

[0142] Step S21, extracting target patches directly adjacent to the boundary of the inner ring according to the topological connection relationship of the boundary.

[0143] In this embodiment, the topological connection relationship refers to the adjacency relationship of geometric elements (edges, faces); and the target patches are the patches directly connected to the inner ring boundary.

[0144] As an optional implementation, the terminal device constructs a double-linked list of inner ring boundary edges, traverses each boundary edge, and obtains the patches adjacent to the edge through the topological relationship, and marks these patches as target patches.

[0145] As another optional implementation, a hash table is used to store the mapping relationship between the boundary edges and the adjacent patches, and the corresponding target patch set can be queried by inputting the inner ring boundary edge ID.

[0146] Step S22, constructing an initial feature surface set according to the target patches.

[0147] In this embodiment, the initial feature surface set is a basic feature surface group composed of target patches.

[0148] The terminal device stores the target facet extracted from the foregoing content into a dynamic array, forms an initial feature facet set initialFeatureFaces after deduplication, and records an adjacency relationship pointer for each facet.

[0149] In step S23, based on the geometric continuity and topological connectivity of the initial feature facet set, an extended facet is iteratively searched to take the initial feature facet and the extended facet as the connection feature, where the initial feature facet and the extended facet are the feature facets.

[0150] In this embodiment, the geometric continuity is judged by the facet normal angle and the boundary curvature, the topological connectivity is judged by the adjacency relationship, and the extended facet is a continuously connected facet outside the initial feature facet.

[0151] As an optional implementation, the terminal device uses a breadth-first search (BFS) algorithm, adds initialFeatureFaces to a queue, and marks them as visited; takes out the facets in the queue, traverses the adjacent facets, calculates the normal angle and the boundary curvature, and if the geometric continuity (such as the normal angle < 10° and the curvature change < a threshold value) is met and not visited, adds the extended facet set extendedFaces and marks it as visited; repeats the foregoing steps until the queue is empty, and finally merges the initial feature facet and the extended facet as the connection feature.

[0152] As another optional implementation, a depth-first search (DFS) is used in combination with geometric constraints, the initial feature facet is taken as a starting point, the adjacent facets are recursively searched, and only when the geometric continuity is met, the extended facet is included to avoid invalid traversal.

[0153] For example, in a three-dimensional model of a mechanical part, there is a circular inner ring on a reference surface. The cylindrical surface adjacent to the inner ring boundary is extracted as a target facet through topological connection relationship, and an initial feature facet set is constructed; then based on the normal angle continuity, the BFS is used to search the chamfer facet smoothly connected to the cylindrical surface as an extended facet, and finally the hole feature including the cylindrical surface and the chamfer facet is determined as a connection feature.

[0154] This embodiment can completely capture the complex feature structure associated with the inner ring by extracting the target facet through the topological connection relationship, iteratively expanding the feature facet in combination with the geometric continuity and the topological connectivity, avoid the misjudgment of design elements caused by the missing feature facet, and provide a complete geometric data basis for subsequent concave-convex analysis and design element classification.

[0155] Based on any of the foregoing embodiments, in the fifth embodiment of the present application, before step S30, the following steps are included:

[0156] Step S301, determining a second normal angle between the average normal of the reference face and the central normal of each feature face in the connection feature.

[0157] In this embodiment, the average normal is obtained by averaging the normal vectors of the reference face vertices; the central normal is obtained by calculating the normal vector at the geometric center of the feature face.

[0158] As an optional implementation, the terminal device traverses all vertices of the reference face, normalizes the sum of the normal vectors of the vertices to obtain the average normal N_base. For each feature face, the geometric center coordinates are calculated, the central normal N_feature is obtained by interpolating the facet normal vectors, and the dot product of the two normals is calculated: dotProduct = N_base · N_feature. The second normal angle is obtained by arccos(dotProduct).

[0159] As another optional implementation, the central normals of all feature faces are calculated in parallel using matrix operations, and the second normal angle set is efficiently calculated by batch dot product operations.

[0160] Step S302, determining the average angle of the connection feature according to the second normal angle.

[0161] In this embodiment, the average angle is the arithmetic mean of all second normal angles.

[0162] As an optional implementation, the terminal device adds all the obtained second normal angle values and divides the sum by the number of feature faces to obtain the average angle meanAngle.

[0163] Step S303, determining the concave-convex property of the connection feature based on the average angle.

[0164] In this embodiment, the concave-convex property is determined by comparing the average angle with a preset threshold.

[0165] As an optional implementation, if meanAngle is less than 90°, the connection feature is determined to be convex; if meanAngle is greater than 90°, the connection feature is determined to be concave; if it is close to 180°, it may be a penetrating feature.

[0166] As another optional implementation, multiple threshold intervals are set (for example, [0°, 75°) for strong convex, [75°, 105°] for plane, and (105°, 180°] for concave), and a more detailed concave-convex type is determined according to the interval in which meanAngle is located.

[0167] As still another optional implementation, if the average angle of the feature face and the reference face is less than 180 degrees, it is a convex feature. If the average angle is greater than 180 degrees, it is a concave feature.

[0168] For example, in a model of an automobile cover mold, a certain reference surface has a circular inner ring, and the connecting feature includes a cylindrical surface and a chamfer surface. The average second normal angle between the average normal of the reference surface and the normal of the center of each feature surface is 102°, which determines that the connecting feature is a recessed feature, corresponding to a punching design element on the mold.

[0169] In this embodiment, the normal angle analysis converts the geometric topological relationship into a quantifiable concave-convex index, providing a clear basis for subsequent design element classification. This method avoids the limitations of traditional depth-based judgment and can accurately identify convex / concave features on complex surfaces, improving the robustness of automatic design element recognition.

[0170] Based on any of the above embodiments, in the sixth embodiment of the present application, step S30 includes:

[0171] Step S41, determining the geometric profile of the connecting feature based on the geometric parameters of the connecting feature.

[0172] In this embodiment, the geometric parameters are calculated based on the vertex coordinates, normal vectors, and topological relationships of the feature surfaces.

[0173] As an optional implementation, the terminal device extracts all boundary edges of the connecting feature to construct a boundary edge set. By calculating the coplanarity and continuity of adjacent boundary edges, the boundary edges are grouped into different profile rings. For each profile ring, a geometric curve (such as a straight line, a circular arc, or a B-spline curve) is fitted to generate a parameterized geometric profile description.

[0174] As another optional implementation, the principal component analysis (PCA) is used to calculate the axial direction of the connecting feature, the feature surfaces are projected onto the axial plane, and a two-dimensional geometric profile is generated by a contour extraction algorithm (such as MarchingSquares), and then mapped back to the three-dimensional space.

[0175] Step S42, identifying the design element in the model based on the concave-convexity and the geometric profile.

[0176] In this embodiment, the concave-convexity and the geometric profile together determine the type of the design element.

[0177] As an optional implementation, a design element knowledge base is established, containing the concave-convexity and geometric profile templates of various elements (such as a circular recess corresponding to a hole, and a rectangular protrusion corresponding to a boss). The calculated concave-convexity and geometric profile are matched with the knowledge base, and if the similarity exceeds a threshold, the corresponding design element is identified.

[0178] As another optional implementation, a machine learning classifier is adopted, the concave-convex parameter and the geometric contour feature vector are input, and the design element type is output. The classifier is optimized by training samples to improve the recognition accuracy of complex geometric features.

[0179] For example, in a mechanical part model, the concave-convex determination of a certain connection feature is concave, the geometric contour fitting is circular, the diameter is 10 mm, and the depth is 5 mm. After matching with the template of "hole" in the knowledge base, the connection feature is identified as a "through hole" design element, and its size parameters are labeled.

[0180] Optionally, after determining the design element corresponding to the reference surface, a three-dimensional bounding box of the design element is calculated, a size threshold range is obtained, and if the three-dimensional bounding box does not satisfy the size threshold range, the design element is deleted. The size range can be user-defined or associated with the current three-dimensional model of the industrial scene.

[0181] In this embodiment, the three-dimensional bounding box is determined by the extreme values of all vertex coordinates of the design element.

[0182] As an optional implementation, the terminal device traverses all vertices of the design element, records the minimum and maximum values (x_min, y_min, z_min) and (x_max, y_max, z_max) of the vertex coordinates, and constructs six faces of the three-dimensional bounding box based on this.

[0183] As another optional implementation, the hierarchical bounding box technology is adopted, the design element is first divided into multiple sub-regions, the bounding boxes of the sub-regions are calculated respectively, and then the overall three-dimensional bounding box is obtained by merging, thereby improving the calculation efficiency.

[0184] In this embodiment, the size threshold range includes the minimum and maximum values of length, width, and height.

[0185] As an optional implementation, the terminal device reads the size threshold parameters input by the user on the system interface, such as the minimum length of 1 mm and the maximum length of 50 mm, and constructs a size threshold range data structure.

[0186] As another optional implementation, according to the industrial scene (such as aerospace, automobile manufacturing) to which the current three-dimensional model belongs, the corresponding size threshold range is automatically matched from a preset scene-size mapping table. For example, in the aerospace scene, the minimum size is set to 0.5 mm and the maximum size is set to 30 mm.

[0187] In this embodiment, the judgment condition is whether the length, width, and height of the three-dimensional bounding box are within the threshold range.

[0188] As an optional implementation, the terminal device compares the length (x_max-x_min), width (y_max-y_min), and height (z_max-z_min) of the three-dimensional bounding box with the minimum and maximum values of the size threshold range, respectively. If the size of any dimension exceeds the threshold range, the element is deleted from the design element set. In the size filtering, the three-dimensional bounding box (BoundingBox) length, width, and height and volume information of all candidate features are calculated, and the user can customize the threshold range (such as minimum depth > 0.5 mm, maximum width < 20 mm, etc.), and the structures that do not meet the expectations are removed.

[0189] For example, in a three-dimensional model of an automobile part, a design element is identified as a boss, and the size of its three-dimensional bounding box is calculated as length 2 mm, width 3 mm, and height 15 mm. Since the size threshold range corresponding to the current industrial scene is length 1-10 mm, width 1-10 mm, and height 1-10 mm, the height of the boss exceeds the threshold, and it is deleted from the design element list.

[0190] Referring to Figure 3 , Figure 3 An example of non-structural functional feature recognition is shown, and unified recognition of bosses, holes, and logos is performed. Figure 3 The bosses, holes, and logos are highlighted in yellow. Referring to Figure 4 , Figure 4 Another example of non-structural functional feature recognition is shown. Optionally, the user can select the highlighted design elements, i.e., classification recognition, and only recognize the logo. Figure 4 In the design element, only the logo is highlighted. Referring to Figure 5 , Figure 5 Another example of non-structural functional feature recognition is shown. Figure 5 The recognition result of the LOGO across the multi-face inlay is shown.

[0191] This embodiment effectively removes noise features or invalid structures that do not meet engineering requirements by calculating the three-dimensional bounding box and combining the size threshold range to filter design elements. This filtering mechanism not only supports user-defined standards, but also adapts to different industrial scene requirements, improving the usability and engineering accuracy of non-structural functional feature recognition results, and providing a more reliable data foundation for subsequent model analysis and optimization. Through geometric contour extraction and convexity analysis, abstract topological relationships are converted into specific engineering semantics, realizing the mapping from geometric models to design intent. This method not only identifies standard features, but also processes custom features through parameterization description, improving the flexibility and engineering practicality of non-structural functional feature recognition.

[0192] Based on any of the above embodiments, referring to Figure 6 , Figure 6For the overall flowchart of the non-structural functional feature recognition method, after determining the geometric model, all the faces of the geometric model are traversed, the inner ring detection of the reference face is performed, if the reference face does not contain an inner ring, the remaining faces are continuously traversed. If the reference face contains an inner ring, typically, such as a flat surface or a weak curved surface, a hole, a logo or other local features are made on it, forming one or more inner ring structures, then the inner ring structure is extracted, the topological expansion is performed on each inner ring, the face set directly connected with the inner ring is preliminarily identified, and the initial feature face forming the connecting feature is extracted. Based on the initial feature face, further according to the geometric continuity (such as the face normal angle, the common edge relationship, the smooth connection) and the topological connectivity, the breadth-first search method is used to iteratively search all the faces continuously connected with the initial feature face, the complete feature face set is constructed, and the entire connecting feature is accurately extracted. The geometric size analysis, such as the feature height, width, depth, area or bounding box size, etc., is performed on the identified connecting feature, and is compared with the feature size range specified by the user, and the features not meeting the requirements are filtered out. This step ensures the controllability and adjustability of the recognition result, and supports the user to make personalized settings according to the simulation requirements. For the boundary between the connecting feature and the reference face, the concave-convex property (for example, through the normal angle, the face angle or the local curvature of the edge) is judged, so as to distinguish whether the feature is concave or convex. If it is a smooth connection (such as a round corner transition), the position relationship of the feature relative to the reference face is judged by using the geometric bounding box position, to assist the concave-convex discrimination. If it is a convex feature, further classification is performed according to the geometric shape, such as: the cylindrical feature can be identified as a boss; the slender column feature can be identified as a rib plate; the complex profile and the text profile feature can be identified as a logo. If it is a concave feature, whether it penetrates the reference face is further judged: if it penetrates the geometric model, it is a through hole; if it does not penetrate, it is judged as a blind hole or a groove according to the profile shape, and can be further distinguished as a circular or a special-shaped type. Then the identified design feature is highlighted in the model.

[0193] Through the above embodiments, multiple feature types in the design model are effectively processed, the recognition of complex nested and multi-level combined features is supported, and good universality and expansibility are possessed. The method is not only suitable for feature simplification processing of the CAD model, but also can be widely applied to scenes such as reverse engineering and geometric semantic modeling.

[0194] The application provides a non-structural functional feature recognition device, which comprises at least one processor and a memory in communication connection with the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the non-structural functional feature recognition method in the above embodiment one.

[0195] Reference will be made to the following description Figure 7, which shows a schematic diagram of the system structure of a non-structural functional feature recognition device suitable for implementing the embodiments of the present application. The non-structural functional feature recognition device in the embodiments of the present application may include, but is not limited to, mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, personal digital assistants (PDAs), tablet computers, and vehicle-mounted terminals, as well as fixed terminals such as digital TVs and desktop computers. Figure 7 The non-structural functional feature recognition device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0196] like Figure 7 As shown, the non-structural functional feature recognition device may include a processing device 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes based on a program stored in a read-only memory (ROM) 1002 or a program loaded from a storage device 1003 into a random access memory (RAM) 1004. The RAM 1004 also stores various programs and data required for the operation of the non-structural functional feature recognition device. The processing device 1001, the ROM 1002, and the RAM 1004 are connected to each other via a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems can be connected to the I / O interface 1006: input devices 1007 including, for example, a touch screen, touchpad, keyboard, mouse, image sensor, microphone, accelerometer, gyroscope, etc.; output devices 1008 including, for example, a liquid crystal display (LCD), speaker, vibrator, etc.; storage devices 1003 including, for example, a magnetic tape, hard disk, etc.; and communication devices 1009. The communication devices 1009 can allow the non-structural functional feature recognition device to communicate with other devices wirelessly or by wire to exchange data. Although the figure shows a non-structural functional feature recognition device with various systems, it should be understood that it is not required to implement or have all of the systems shown. More or fewer systems may alternatively be implemented or have.

[0197] In particular, according to embodiments of the present application, the processes described above with reference to the flowcharts can be implemented as a computer software program. For example, embodiments of the present application include a computer program product comprising a computer program carried on a computer readable medium, the computer program comprising program code for performing the methods illustrated by the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from a storage device 1003, or installed from a read-only memory 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the methods of the embodiments of the present application are performed.

[0198] The non-structural functional feature recognition device provided by the present application adopts the non-structural functional feature recognition method in the above-mentioned embodiments, and can solve the technical problem that the related art is sensitive to geometric shapes and sizes, resulting in insufficient robustness of feature recognition. Compared with the prior art, the non-structural functional feature recognition device provided by the present application has the same beneficial effects as the non-structural functional feature recognition device provided by the above-mentioned embodiments, and other technical features in the non-structural functional feature recognition device are the same as the features disclosed in the previous embodiment method, which will not be repeated here.

[0199] It should be understood that parts of the present application can be realized by hardware, software, firmware or a combination thereof. In the description of the above-mentioned embodiments, specific features, structures, materials or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0200] The above is merely specific embodiments of the present application, but the protection scope of the present application is not limited thereto, and any person skilled in the art can easily think of changes or replacements within the technical scope disclosed by the present application, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

[0201] The present application provides a computer readable storage medium having stored thereon computer readable program instructions (i.e. computer program) for performing the non-structural functional feature recognition method in the above-mentioned embodiments.

[0202] The computer readable storage medium provided in the application may, for example, be a U disk, but is not limited to an electric, magnetic, optical, electromagnetic, infrared, or semiconductor system, system, or device, or any combination of the above. More specific examples of the computer readable storage medium may include, but are not limited to, an electric connection with one or more conductive wires, a portable computer disk, a hard disk, a random access memory (RAM), a read only memory (ROM), an erasable programmable read only memory (EPROM), a flash memory, an optical fiber, a portable compact disk read only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the embodiment, the computer readable storage medium may be any tangible medium containing or storing a program, which can be used by or in combination with an instruction execution system, system, or device. The program code contained on the computer readable storage medium can be transmitted by any suitable medium, including but not limited to an electric wire, an optical cable, a radio frequency (RF), or any suitable combination of the above.

[0203] The computer readable storage medium described above may be contained in a non-structural functional feature recognition device, or may exist separately without being assembled into the non-structural functional feature recognition device.

[0204] The computer readable storage medium described above carries one or more programs, which, when executed by the non-structural functional feature recognition device, cause the non-structural functional feature recognition device to: determine a reference surface currently traversed, and determine whether the reference surface has an inner ring based on the topological nature of the reference surface.

[0205] If the reference surface has an inner ring, search for a feature surface associated with the inner ring to determine a connecting feature.

[0206] Determine a design element corresponding to the reference surface based on the concave-convex nature of the connecting feature.

[0207] Computer program code for carrying out operations of the present application can be written in any combination of one or more programming languages, including an object oriented programming language such as Java, Smalltalk, C++ or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code can execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the latter scenario, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computer (for example, through the Internet using an Internet Service Provider).

[0208] The flow diagrams and the block diagrams in the drawings are illustrations of architectures, functionalities, and operations of possible implementations of systems, methods, and computer program products according to various embodiments of present application. In this regard, each block in the flow diagrams or block diagrams can represent a module, a segment, or a portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that in some alternative implementations, the functions noted in the blocks can occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently or the blocks may

[0209] The modules involved in the embodiments of the present application can be implemented in the form of software or in the form of hardware. In some cases, the name of the module does not constitute a limitation on the module itself.

[0210] The readable storage medium provided by the present application is a computer readable storage medium, which stores computer readable program instructions (i.e., computer programs) for executing the above-mentioned non-structural functional feature recognition method, and can solve the technical problem that the related art is sensitive to geometric shape and size, resulting in insufficient robustness of feature recognition. Compared with the prior art, the computer readable storage medium provided by the present application has the same beneficial effects as the non-structural functional feature recognition method provided by the above-mentioned embodiments, and will not be described here.

[0211] The embodiment of the present application provides a computer program product, comprising a computer program, which realizes the steps of the non-structural functional feature identification method when executed by a processor.

[0212] The computer program product provided by the present application can solve the technical problem of insufficient feature recognition robustness caused by strong sensitivity to geometric shapes and sizes in the related art. Compared with the prior art, the beneficial effects of the computer program product provided by the embodiment of the present application are the same as those of the non-structural functional feature identification method provided by the above-mentioned embodiment, and are not described here.

[0213] The above is only the preferred embodiment of the present application, and does not limit the patent scope of the present application. Any equivalent structure or equivalent process transformation using the content of the specification and drawings, or direct or indirect application in other related technical fields, is also included in the patent processing scope of the present application.

Claims

1. A method for identifying non-structural functional features, characterized in that: The non-structural functional feature recognition method comprises: Determine the currently traversed reference plane, and determine whether the reference plane has an inner ring based on the topology of the reference plane; If the reference plane has an inner ring, extracting a target face directly adjacent to the boundary according to the topological connection relationship of the boundary of the inner ring; Constructing an initial feature face set according to the target face; Iteratively searching for extended facets based on the geometric continuity and topological connectivity of the initial feature face set, so as to use the initial feature facets and the extended facets as connection features, wherein the initial feature facets and the extended facets serve as the feature faces; Determine a second normal angle between an average normal of the reference surface in the connection feature and a center normal of each feature surface; Determining a mean angle of the connection features according to the second normal angle; determining the concavity and convexity of the connection feature based on the mean value of the angle; The design element corresponding to the reference surface is determined based on the concavity and convexity of the connection feature.

2. The non-structural functional feature recognition method according to claim 1, wherein: Determining the currently traversed reference plane includes: Determine the currently traversed reference surface based on the patch identifier; or Determine the initial reference surface currently traversed according to the patch identifier; Determining, in a neighborhood of the initial reference plane, a neighboring facet that satisfies a smooth connectivity requirement with the initial reference plane, wherein a first normal angle between the facet in the neighborhood and the initial reference plane is less than a preset angle, and / or a change in boundary curvature between the facet in the neighborhood and the initial reference plane is less than a preset threshold, and determining that the facet is a neighboring facet that satisfies the smooth connectivity requirement; The reference plane is obtained by removing shared internal edges in an extended base plane region formed by the initial reference plane and the neighboring face patches.

3. The non-structural functional feature recognition method according to claim 2, wherein: The removing of shared internal edges in the extended base surface region formed by the initial reference surface and the neighboring facets to obtain the reference surface includes: Determining edges in an extended base surface area formed by the initial reference surface and the neighboring surface patches; Determine isolated edges in the edge lines, and determine the edges other than the isolated edges in the edge lines as shared internal edges; or determine the shared internal edges in the edge lines, and determine the edges other than the shared internal edges in the edge lines as isolated edges; The shared internal edges in the extended base surface area are removed to obtain the reference surface.

4. The non-structural functional feature recognition method according to claim 3, wherein: The determining of isolated edges in the edge line includes: constructing an undirected edge structure based on the hash values ​​of the edges; Counting the number of occurrences of each edge in the undirected edge structure according to a hash table; The target edge line whose number of occurrences is a preset value is determined to be the isolated edge.

5. The non-structural functional feature recognition method according to claim 1, wherein: The determining whether the reference plane has an inner ring based on the topology of the reference plane includes: Performing closed loop search based on isolated edges in the reference plane to determine a closed boundary loop; If the number of the closed boundary rings is greater than two, it is determined that an inner ring exists on the reference plane.

6. The non-structural functional feature recognition method according to claim 1, wherein: The determining of the design element corresponding to the reference surface based on the concavity and convexity of the connection feature includes: determining a geometric profile of the connection feature based on geometric parameters of the connection feature; The design element is identified in a model based on the concavity and convexity and the geometric profile.

7. A non-structural functional feature recognition device, characterized in that: The non-structural functional feature recognition device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of the non-structural functional feature recognition method according to any one of claims 1 to 6.

8. A storage medium, characterized in that: The storage medium is a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the steps of the non-structural functional feature recognition method according to any one of claims 1 to 6 are implemented.

Images