Data processing method and device for process feature recognition

Abstract

The invention discloses a data processing method and device for process feature recognition. To-be-processed data is obtained, the to-be-processed data is used for representing data of a to-be-processed workpiece, and the to-be-processed workpiece is a workpiece with to-be-recognized process characteristics; performing extraction processing based on three-dimensional features on the to-be-processed data to obtain to-be-processed three-dimensional feature data, the to-be-processed three-dimensional feature data being data used for representing three-dimensional object features of the to-be-processed workpiece; the to-be-processed three-dimensional feature data are subjected to process feature prediction processing based on a preset process feature prediction model, target process feature data are obtained, the preset process feature prediction model is a graph neural network model, and the target process feature data are data used for representing process features obtained through prediction of the to-be-processed workpiece. The three-dimensional feature data of the to-be-processed workpiece is subjected to process feature prediction processing through the preset process feature prediction model based on the graph neural network model, and the technical effect of improving the process feature recognition accuracy is achieved.

Description

Technical Field

[0001] The present application relates to the field of computers, and in particular, to a data processing method and device for process feature identification. Background Art

[0002] Process feature recognition technology usually adopts traditional topology-based methods, such as inferring the location of process features based on the vertical and overlapping relationships between faces and edges.

[0003] In the prior art, the classification of edges or faces of three-dimensional objects through machine learning may be based on point clouds or images sampled from multiple camera positions. After obtaining the point cloud image of the workpiece to be inspected, a point cloud three-dimensional model of the workpiece is constructed based on the point cloud image, and finally the process features are recognized. However, due to factors such as noise and distortion, there is a large error in determining the precise position of process features based on point clouds.

[0004] Therefore, the process feature recognition in the prior art has the problem of low accuracy. Summary of the invention

[0005] The main purpose of this application is to provide a data processing method and device for process feature identification, so as to solve the problem of low accuracy of process feature identification in the prior art and achieve the technical effect of improving the accuracy of process feature identification.

[0006] In order to achieve the above-mentioned purpose, in a first aspect of the present application, a data processing method for process feature identification is proposed, comprising:

[0007] Acquire data to be processed, wherein the data to be processed is data used to represent a workpiece to be processed, and the workpiece to be processed is a workpiece whose process features are to be identified;

[0008] Performing extraction processing based on three-dimensional features on the data to be processed to obtain three-dimensional feature data to be processed, wherein the three-dimensional feature data to be processed is data used to represent the three-dimensional object features of the workpiece to be processed;

[0009] The three-dimensional feature data to be processed is subjected to process feature prediction processing based on a preset process feature prediction model to obtain target process feature data, wherein the preset process feature prediction model is a graph neural network model, and the target process feature data is data used to represent the process features predicted for the workpiece to be processed.

[0010] Furthermore, the three-dimensional feature-based extraction processing is performed on the data to be processed, and the three-dimensional feature data to be processed is obtained, including:

[0011] Performing mapping processing based on the geometric elements of the object on the data to be processed to obtain preprocessed data, wherein the preprocessed data is data after the geometric elements of the workpiece are mapped and used to represent the data to be processed;

[0012] Performing feature extraction processing based on the first geometric element on the preprocessed data to obtain first geometric element feature data, wherein the first geometric element feature data is surface feature data for representing the workpiece to be processed;

[0013] Performing feature extraction processing based on a second geometric element on the preprocessed data to obtain second geometric element feature data, wherein the second geometric element feature data is edge feature data for representing a workpiece to be processed;

[0014] Performing graph construction processing on the preprocessed data to obtain constructed graph data;

[0015] The three-dimensional feature data to be processed is obtained according to the first geometric element feature data, the second geometric element feature data and the construction graph data.

[0016] Further, the preprocessed data is subjected to feature extraction processing based on the first geometric element to obtain feature data of the first geometric element, including:

[0017] Performing sampling processing based on a first geometric element on the preprocessed data to obtain first sampling data, wherein the first sampling data is sampling point data for representing a surface of a workpiece to be processed;

[0018] Perform feature extraction processing on the first sampling data to obtain the first geometric element feature data.

[0019] Further, the preprocessed data is subjected to graph construction processing to obtain constructed graph data including:

[0020] Performing recognition processing based on adjacent graphic elements on the preprocessed data to obtain adjacent graphic element data, wherein the adjacent graphic element data is data used to represent adjacent graphic elements in the workpiece to be processed;

[0021] The adjacent graph metadata is processed for graph data construction to obtain constructed graph data.

[0022] Furthermore, before obtaining the data to be processed, the method further includes:

[0023] Acquire sample three-dimensional object data, wherein the sample three-dimensional object data is relevant data used to represent the sample three-dimensional object;

[0024] Performing extraction processing based on three-dimensional features on the sample three-dimensional object data to obtain sample three-dimensional feature data;

[0025] The preset graph neural network model is trained according to the sample three-dimensional feature data to obtain the preset process feature prediction model.

[0026] Furthermore, the preset graph neural network model is trained according to the sample three-dimensional feature data to obtain the preset process feature prediction model, which includes:

[0027] Performing model training processing on a preset graph neural network model according to the sample three-dimensional feature data to obtain a process feature prediction model;

[0028] Performing a model test process based on a test sample on the process technology feature prediction model to obtain test feature data, wherein the test feature data is feature data used to represent the test accuracy of the process technology feature prediction model;

[0029] Compare the accuracy value corresponding to the test feature data with the preset accuracy threshold,

[0030] If the accuracy value corresponding to the test feature data is greater than or equal to the preset accuracy threshold, the preset process feature prediction model is obtained, wherein the preset process feature prediction model is the process feature prediction model;

[0031] If the accuracy value corresponding to the test feature data is less than the preset accuracy threshold, the process feature prediction model is iteratively trained until the accuracy value corresponding to the test feature data is greater than or equal to the preset accuracy threshold, so as to obtain the preset process feature prediction model.

[0032] According to a second aspect of the present application, a data processing device for process feature recognition is proposed, comprising:

[0033] A data acquisition module, used for acquiring data to be processed, wherein the data to be processed is data used to represent a workpiece to be processed, and the workpiece to be processed is a workpiece whose process features are to be identified;

[0034] A feature extraction module, used for performing extraction processing based on three-dimensional features on the data to be processed to obtain three-dimensional feature data to be processed, wherein the three-dimensional feature data to be processed is data used to represent the three-dimensional object features of the workpiece to be processed;

[0035] A process feature prediction module is used to perform process feature prediction processing on the three-dimensional feature data to be processed based on a preset process feature prediction model to obtain target process feature data, wherein the preset process feature prediction model is a graph neural network model, and the target process feature data is data used to represent the process features predicted for the workpiece to be processed.

[0036] Furthermore, the feature extraction module includes:

[0037] A mapping module, used for performing mapping processing on the data to be processed based on the geometric elements of the object to obtain pre-processed data, wherein the pre-processed data is data used to represent the mapped geometric elements of the workpiece of the data to be processed;

[0038] A first feature extraction module, used for performing feature extraction processing based on a first geometric element on the preprocessed data to obtain first geometric element feature data, wherein the first geometric element feature data is surface feature data for representing a workpiece to be processed;

[0039] A second feature extraction module, used for performing feature extraction processing based on a second geometric element on the preprocessed data to obtain second geometric element feature data, wherein the second geometric element feature data is edge feature data for representing the workpiece to be processed;

[0040] A graph construction module, used to perform graph construction processing on the preprocessed data to obtain constructed graph data;

[0041] A three-dimensional feature extraction module is used to obtain the three-dimensional feature data to be processed according to the first geometric element feature data, the second geometric element feature data and the construction image data.

[0042] According to a third aspect of the present application, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable the computer to execute the above-mentioned data processing method for process feature identification.

[0043] According to the fourth aspect of the present application, an electronic device is proposed, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor executes the above-mentioned data processing method for process feature identification.

[0044] The technical solution provided by the embodiments of the present application may have the following beneficial effects:

[0045] In the present application, data to be processed is obtained, wherein the data to be processed is data used to represent the workpiece to be processed, and the workpiece to be processed is a workpiece whose process features are to be identified; the data to be processed is subjected to extraction processing based on three-dimensional features to obtain three-dimensional feature data to be processed, wherein the three-dimensional feature data to be processed is data used to represent the three-dimensional object features of the workpiece to be processed; the three-dimensional feature data to be processed is subjected to process feature prediction processing based on a preset process feature prediction model to obtain target process feature data, wherein the preset process feature prediction model is a graph neural network model, and the target process feature data is data used to represent the process features predicted for the workpiece to be processed. Process feature prediction processing is performed on the three-dimensional feature data of the workpiece to be processed by using a preset process feature prediction model based on a graph neural network model, thereby solving the problem of low accuracy in process feature recognition in the prior art and achieving the technical effect of improving the accuracy of process feature recognition. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] The drawings constituting a part of this application are used to provide a further understanding of this application, so that other features, purposes and advantages of this application become more obvious. The schematic embodiment drawings and their descriptions of this application are used to explain this application and do not constitute an improper limitation on this application. In the drawings:

[0047] Figure 1 A flowchart of a data processing method for process feature identification provided in this application;

[0048] Figure 2 A flowchart of a data processing method for process feature identification provided in this application;

[0049] Figure 3 A flowchart of a data processing method for process feature identification provided in this application;

[0050] Figure 4 A schematic diagram of a data processing device for process feature identification provided by the present application;

[0051] Figure 5 A schematic diagram of another data processing device for process feature identification provided in the present application. DETAILED DESCRIPTION

[0052] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.

[0053] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequential order. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present application described here. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0054] In the present application, the terms "upper", "lower", "left", "right", "front", "back", "top", "bottom", "inner", "outer", "middle", "vertical", "horizontal", "lateral", "longitudinal" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the drawings. These terms are mainly used to better describe the present application and its embodiments, and are not used to limit the indicated devices, elements or components to have a specific orientation, or to be constructed and operated in a specific orientation.

[0055] In addition, some of the above terms may be used to express other meanings in addition to indicating orientation or positional relationship. For example, the term "on" may also be used to express a certain dependency or connection relationship in some cases. For those of ordinary skill in the art, the specific meanings of these terms in this application can be understood according to specific circumstances.

[0056] In addition, the terms "installed", "set", "provided with", "connected", "connected", and "socketed" should be understood in a broad sense. For example, "connection" can be a fixed connection, a detachable connection, or an integral structure; it can be a mechanical connection or an electrical connection; it can be a direct connection, or an indirect connection through an intermediate medium, or it can be an internal connection between two devices, elements, or components. For those of ordinary skill in the art, the specific meanings of the above terms in this application can be understood according to specific circumstances.

[0057] When processing a workpiece, it is usually necessary to obtain the process features of the workpiece so as to perform corresponding processing operations according to the process features, such as the features of the workpiece processing technology such as the workpiece weld and workpiece surface. In the prior art, the process feature recognition technology usually adopts the traditional topology-based approach, such as inferring the location of the process features based on the vertical and coincident relationships between faces and edges.

[0058] In the prior art, the edge or face classification of three-dimensional objects through machine learning may be based on point clouds or pictures obtained by sampling from multiple camera positions, and process feature recognition is performed based on point clouds. After obtaining the point cloud image of the workpiece to be inspected, a point cloud three-dimensional model of the workpiece is constructed based on the point cloud image, and finally the process feature recognition is realized. However, due to the influence of factors such as noise and distortion, there is a large error in determining the precise position of the process feature based on the point cloud, and the process feature recognition has a low accuracy rate.

[0059] In some optional embodiments of the present application, a data processing method for process feature recognition is proposed. Figure 1 A flowchart of a data processing method for process feature identification provided in this application, such as Figure 1 As shown, the method comprises the following steps:

[0060] S101: Obtain data to be processed;

[0061] The data to be processed is data used to represent the workpiece to be processed. The workpiece to be processed is a workpiece whose process features are to be identified. The data to be processed is file data used to represent the three-dimensional model of the workpiece to be processed. Different CAD software have different formats. The file formats of the three-dimensional model file data of the workpiece to be processed are different. The three-dimensional model files of the workpiece to be processed in different formats are format converted to obtain three-dimensional model file data in a common format. For example, the common format may be a STEP file. After obtaining the three-dimensional model file data of the workpiece to be processed, the three-dimensional model file data is converted into the common format STEP to obtain the file to be processed.

[0062] S102: performing extraction processing based on three-dimensional features on the data to be processed to obtain three-dimensional feature data to be processed;

[0063] The three-dimensional feature data to be processed is data used to represent the three-dimensional object features of the workpiece to be processed. The three-dimensional feature data to be processed includes surface feature data, edge feature data and construction graph data, so as to realize the process feature prediction processing according to the three-dimensional feature data to be processed. The three-dimensional feature data for process feature prediction is obtained by performing three-dimensional feature extraction processing on the data to be processed, and the extracted three-dimensional features to be processed are used as the input of the preset process feature prediction model to realize the process feature recognition prediction of the workpiece to be identified.

[0064] In some optional embodiments of the present application, a data processing method for process feature recognition is proposed. Figure 2 A flowchart of a data processing method for process feature identification provided in this application, such as Figure 1 As shown, the method comprises the following steps:

[0065] S201: performing mapping processing on the data to be processed based on the geometric elements of the object to obtain pre-processed data;

[0066] The preprocessed data is the data after the geometric elements of the workpiece to be processed are mapped. The geometric elements of the workpiece to be processed are traversed. The geometric elements include faces and edges. The faces and edges of the workpiece to be processed are mapped to a two-dimensional image (hereinafter referred to as UV) coordinate system to obtain parametric equations corresponding to the geometric elements. The face of the workpiece to be processed is mapped to the UV coordinate system to obtain the parametric equation of the face of the workpiece to be processed; the edge of the workpiece to be processed is mapped to the UV coordinate system to obtain the parametric equation of the edge of the workpiece to be processed.

[0067] S202: performing feature extraction processing based on the first geometric element on the preprocessed data to obtain first geometric element feature data;

[0068] The first geometric element feature data is surface feature data used to represent the workpiece to be processed.

[0069] In some optional embodiments of the present application, a data processing method for process feature identification is proposed, the method comprising: performing sampling processing on preprocessed data based on a first geometric element to obtain first sampling data, wherein the first sampling data is sampling point data for representing the surface of a workpiece to be processed; performing feature extraction processing on the first sampling data to obtain first geometric element feature data.

[0070] The parametric equation of the surface of the workpiece to be processed in the preprocessing data is processed to obtain the maximum value of the parametric equation, and the maximum value of the parametric equation of the surface is sampled. For example, the maximum value of the parametric equation is taken out and divided into 10 parts for sampling, and 10*10 sampling points are generated on the surface; feature extraction is performed on each sampling point in the surface sampling data to obtain feature data of the sampling point, including: point coordinates, normal vector (pointing to the outside of the surface), curvature of the point and other feature data.

[0071] S203: performing feature extraction processing based on the second geometric element on the preprocessed data to obtain second geometric element feature data;

[0072] The second geometric element feature data is edge feature data used to represent the workpiece to be processed.

[0073] In some optional embodiments of the present application, a data processing method for process feature identification is proposed, the method comprising: performing sampling processing on preprocessed data based on a first geometric element to obtain first sampling data, wherein the first sampling data is sampling point data for representing the surface of a workpiece to be processed; performing feature extraction processing on the first sampling data to obtain first geometric element feature data.

[0074] The parametric equation of the edge of the workpiece to be processed in the preprocessing data is subjected to maximum value processing of the parametric equation, the maximum value of the parametric equation of the edge is taken out, and the maximum value of the parametric equation of the edge is sampled, such as taking out the maximum value of the parametric equation and dividing it into 10 parts for sampling, and generating 10 sampling points on the edge; feature extraction processing is performed on each feature point in the edge sampling data to obtain feature data of the sampling point, and curvature feature data of each feature point is extracted. Edge feature extraction is performed on the edge to obtain edge feature data such as edge coordinates and tangent vectors, and second geometric element feature data is obtained according to the feature data of the edge sampling point and the edge feature data.

[0075] S204: Perform graph construction processing on the preprocessed data to obtain constructed graph data;

[0076] In some optional embodiments of the present application, a data processing method for process feature recognition is proposed, the method comprising:

[0077] The object is identified based on geometric elements to obtain object geometric element data, which includes object surface data and object edge data; a three-dimensional object construction graph is processed according to the object surface data and object edge data; graph node data is generated from the object surface data, and graph change data is generated from the object point data, with the object surface as the graph node and the object edge as the graph edge; a face adjacency judgment is performed on the object surface data, and if two faces of the object are adjacent, an edge will be generated geometrically. At the same time, an edge is added between the two corresponding nodes in the graph to construct a graph in which the object surface is converted to a node, and process construction graph data is obtained; a graph update processing based on edge conversion is performed on the process construction graph data, and the edge data in the process construction graph data is converted to generate graph node data of a new graph; by performing an adjacency judgment on the edges in the process construction graph data, and by checking whether there is node sharing between the graph nodes corresponding to the two object edges respectively, a new edge is added between the two nodes of the new graph corresponding to the two object edges respectively. Take the edges of the object as nodes. When the edges of the object are coplanar, add an edge to the corresponding node in the graph to construct a graph in which the edges of the object are converted into nodes.

[0078] The preprocessing data is processed based on adjacent graphic elements to obtain adjacent graphic element data, wherein the adjacent graphic element data is data used to represent adjacent graphic elements in the workpiece to be processed; and the adjacent graphic element data is processed to construct graphic data to obtain constructed graphic data.

[0079] In some optional embodiments of the present application, a data processing method for process feature recognition is proposed, the method comprising:

[0080] In the welding scenario, two objects are separated to reserve a certain space for welding. However, two topologically independent objects will form two unconnected subgraphs, which is a big obstacle in the graph neural network, hindering the message transmission through neighbor nodes, and making it difficult for the graph neural network to learn the relationship between objects.

[0081] The object data is traversed based on geometric elements to obtain object geometric element data, which includes object surface data. The object surface data is screened based on distance features to obtain welding surface data, wherein the welding surface data are the two closest surfaces on different objects, the distance between the object surfaces is the minimum distance between the closest points on the two surfaces, and the welding surface is used to represent two surfaces on different objects that are temporarily separated due to welding. A three-dimensional object construction graph is processed based on the welding surface data, and graph node data is generated for the edges of the two surfaces in the welding surface data, and edge data is generated for the graph nodes corresponding to the two welding surfaces, that is, an edge is connected between the graph nodes corresponding to the two welding surfaces, and a graph is constructed in which adjacent object edges are converted into nodes.

[0082] For example, for two objects to be welded, a traversal process based on geometric elements is performed on the two objects to be welded, all faces of the two objects are traversed, and the two faces closest to each other are screened based on all faces of the two objects obtained by the traversal, thereby obtaining the welding faces of the two objects to be welded.

[0083] In another optional embodiment, the object edge data is identified and processed based on attribute features to obtain edge attribute feature data, wherein the edge attribute feature data is feature data used to represent the situation where the object edge crosses a plane. A preset graph neural network model is trained based on the identified edge attribute feature data and the above-mentioned three-dimensional feature data to be processed to obtain a preset process feature prediction model. By identifying the attribute features of the edge, the accuracy of model training is improved during the model training process.

[0084] S205: Obtaining three-dimensional feature data to be processed according to the first geometric element feature data, the second geometric element feature data and the construction graph data.

[0085] S103: Performing process feature prediction processing on the three-dimensional feature data to be processed based on a preset process feature prediction model to obtain target process feature data.

[0086] The preset process feature prediction model is a graph neural network model, and the target process feature data is data used to represent the process features predicted for the workpiece to be processed.

[0087] In some optional embodiments of the present application, a data processing method for process feature recognition is proposed. Figure 3A flowchart of a data processing method for process feature identification provided in this application, such as Figure 1 As shown, the method comprises the following steps:

[0088] S301: Acquire sample three-dimensional object data;

[0089] The sample three-dimensional object data is the relevant data used to represent the sample three-dimensional object. The sample three-dimensional object data is the data of the three-dimensional object with marked process features. For three-dimensional object file data in different formats, the three-dimensional object file data is format converted to obtain three-dimensional object model file data in a universal format. For example, the universal format may be a STEP file. The three-dimensional object file data is converted into the universal format STEP to obtain sample three-dimensional object data.

[0090] S302: performing extraction processing based on three-dimensional features on the sample three-dimensional object data to obtain sample three-dimensional feature data;

[0091] In an optional embodiment of the present application, the sample three-dimensional object data is subjected to extraction processing based on three-dimensional features to obtain the sample three-dimensional feature data including:

[0092] The sample three-dimensional object data is subjected to mapping processing based on the object geometric elements to obtain preprocessing data, wherein the preprocessing data is data after the workpiece geometric elements are mapped and used to represent the sample three-dimensional object data; the preprocessing data is subjected to feature extraction processing based on the first geometric element to obtain sample first geometric element feature data, wherein the sample first geometric element feature data is surface feature data used to represent the sample three-dimensional object; the preprocessing data is subjected to feature extraction processing based on the second geometric element to obtain sample second geometric element feature data, wherein the sample second geometric element feature data is edge feature data used to represent the sample three-dimensional object; the preprocessing data is subjected to graph construction processing to obtain sample construction graph data; the sample three-dimensional feature data is obtained according to the sample first geometric element feature data, the sample second geometric element feature data and the sample construction graph data.

[0093] S303: Perform model training processing on the preset graph neural network model according to the sample three-dimensional feature data to obtain a preset process feature prediction model.

[0094] In some optional embodiments of the present application, a data processing method for process feature recognition is proposed, the method comprising the following steps:

[0095] The preset graph neural network model is trained according to the sample three-dimensional feature data to obtain a process feature prediction model; the process feature prediction model is tested based on the test sample to obtain test feature data, wherein the test feature data is feature data used to represent the test accuracy of the process feature prediction model; the accuracy value corresponding to the test feature data is compared with the preset accuracy threshold, and if the accuracy value corresponding to the test feature data is greater than or equal to the preset accuracy threshold, the preset process feature prediction model is obtained, wherein the preset process feature prediction model is the process feature prediction model; if the accuracy value corresponding to the test feature data is less than the preset accuracy threshold, the process feature prediction model is iteratively trained until the accuracy value corresponding to the test feature data is greater than or equal to the preset accuracy threshold, and the preset process feature prediction model is obtained.

[0096] Building a graph neural network model includes compressing the three-dimensional features of the samples extracted above, and constructing the face encoder FaceEncoder and the edge encoder EdgeEncoder respectively. Taking the face encoder as an example, Conv2d can be used as the convolution layer, BatchNorm2d as the normalization layer, LeakyReLU as the activation layer, and Dropout, Linear, etc. can be used to train the preset graph neural network model based on multiple convolution layers, which can be a GINEConv convolution layer combined with an aggregation function. By adjusting the sample data for each training, and adjusting the learning rate (StepLR) and the optimizer at equal intervals to optimize the training of the graph neural network model, a preset process feature prediction model is obtained.

[0097] In some optional embodiments of the present application, a data processing device for process feature recognition is proposed. Figure 4 A schematic diagram of a data processing device for process feature recognition provided in this application includes:

[0098] A data acquisition module 41 is used to acquire data to be processed, wherein the data to be processed is data used to represent a workpiece to be processed, and the workpiece to be processed is a workpiece whose process features are to be identified;

[0099] The feature extraction module 42 is used to perform a three-dimensional feature extraction process on the data to be processed to obtain the three-dimensional feature data to be processed, wherein the three-dimensional feature data to be processed is data used to represent the three-dimensional object features of the workpiece to be processed;

[0100] The process feature prediction module 43 is used to perform process feature prediction processing on the three-dimensional feature data to be processed based on a preset process feature prediction model to obtain target process feature data, wherein the preset process feature prediction model is a graph neural network model, and the target process feature data is data used to represent the process features predicted for the workpiece to be processed.

[0101] In some optional embodiments of the present application, a data processing device for process feature recognition is proposed. Figure 5 A schematic diagram of another data processing device for process feature recognition provided in this application includes:

[0102] A mapping module 51 is used to perform mapping processing on the data to be processed based on the geometric elements of the object to obtain pre-processed data, wherein the pre-processed data is the data after the geometric elements of the workpiece are mapped to represent the data to be processed;

[0103] A first feature extraction module 52, used for performing feature extraction processing based on a first geometric element on the preprocessed data to obtain first geometric element feature data, wherein the first geometric element feature data is surface feature data for representing a workpiece to be processed;

[0104] A second feature extraction module 53 is used to perform feature extraction processing based on a second geometric element on the preprocessed data to obtain second geometric element feature data, wherein the second geometric element feature data is edge feature data for representing the workpiece to be processed;

[0105] A graph construction module 54, used to perform graph construction processing on the preprocessed data to obtain constructed graph data;

[0106] The three-dimensional feature extraction module 55 is used to obtain the three-dimensional feature data to be processed according to the first geometric element feature data, the second geometric element feature data and the construction graph data.

[0107] The specific manner of executing the operation of each unit in the above embodiment has been described in detail in the embodiment of the method, and will not be elaborated here.

[0108] In summary, in the present application, data to be processed is obtained, wherein the data to be processed is data used to represent the workpiece to be processed, and the workpiece to be processed is a workpiece whose process features are to be identified; the data to be processed is subjected to extraction processing based on three-dimensional features to obtain three-dimensional feature data to be processed, wherein the three-dimensional feature data to be processed is data used to represent the three-dimensional object features of the workpiece to be processed; the three-dimensional feature data to be processed is subjected to process feature prediction processing based on a preset process feature prediction model to obtain target process feature data, wherein the preset process feature prediction model is a graph neural network model, and the target process feature data is data used to represent the process features predicted for the workpiece to be processed. Process feature prediction processing is performed on the three-dimensional feature data of the workpiece to be processed by using a preset process feature prediction model based on a graph neural network model, thereby solving the problem of low accuracy in process feature recognition in the prior art and achieving the technical effect of improving the accuracy of process feature recognition.

[0109] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0110] Obviously, those skilled in the art should understand that the above-mentioned units or steps of the present application can be implemented by a general-purpose computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, and optionally, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. In this way, the present application is not limited to any specific combination of hardware and software.

[0111] The above description is only the preferred embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A data processing method for process feature identification, characterized in that: include: Acquire data to be processed, wherein the data to be processed is data used to represent a workpiece to be processed, and the workpiece to be processed is a workpiece whose process features are to be identified; Performing extraction processing based on three-dimensional features on the data to be processed to obtain three-dimensional feature data to be processed, wherein the three-dimensional feature data to be processed is data used to represent the three-dimensional object features of the workpiece to be processed; The three-dimensional feature data to be processed is subjected to process feature prediction processing based on a preset process feature prediction model to obtain target process feature data, wherein the preset process feature prediction model is a graph neural network model, and the target process feature data is data used to represent the process features predicted for the workpiece to be processed.

2. The data processing method according to claim 1, characterized in that: The three-dimensional feature-based extraction processing is performed on the data to be processed, and the three-dimensional feature data to be processed is obtained, including: Performing mapping processing based on the geometric elements of the object on the data to be processed to obtain preprocessed data, wherein the preprocessed data is data after the geometric elements of the workpiece are mapped and used to represent the data to be processed; Performing feature extraction processing based on the first geometric element on the preprocessed data to obtain first geometric element feature data, wherein the first geometric element feature data is surface feature data for representing the workpiece to be processed; Performing feature extraction processing based on a second geometric element on the preprocessed data to obtain second geometric element feature data, wherein the second geometric element feature data is edge feature data for representing a workpiece to be processed; Performing graph construction processing on the preprocessed data to obtain constructed graph data; The three-dimensional feature data to be processed is obtained according to the first geometric element feature data, the second geometric element feature data and the construction graph data.

3. The data processing method according to claim 2, characterized in that: Performing feature extraction processing based on the first geometric element on the preprocessed data to obtain first geometric element feature data includes: Performing sampling processing based on a first geometric element on the preprocessed data to obtain first sampling data, wherein the first sampling data is sampling point data for representing a surface of a workpiece to be processed; Perform feature extraction processing on the first sampling data to obtain the first geometric element feature data.

4. The data processing method according to claim 2, characterized in that: The preprocessed data is subjected to graph construction processing to obtain constructed graph data including: Performing recognition processing based on adjacent graphic elements on the preprocessed data to obtain adjacent graphic element data, wherein the adjacent graphic element data is data used to represent adjacent graphic elements in the workpiece to be processed; The adjacent graph metadata is processed for graph data construction to obtain constructed graph data.

5. The data processing method according to claim 1, characterized in that: Before obtaining the data to be processed, the method further includes: Acquire sample three-dimensional object data, wherein the sample three-dimensional object data is relevant data used to represent the sample three-dimensional object; Performing extraction processing based on three-dimensional features on the sample three-dimensional object data to obtain sample three-dimensional feature data; The preset graph neural network model is trained according to the sample three-dimensional feature data to obtain the preset process feature prediction model.

6. The data processing method according to claim 5, characterized in that: The preset graph neural network model is trained according to the sample three-dimensional feature data to obtain the preset process feature prediction model, including: Performing model training processing on a preset graph neural network model according to the sample three-dimensional feature data to obtain a process feature prediction model; Performing a model test process based on a test sample on the process technology feature prediction model to obtain test feature data, wherein the test feature data is feature data used to represent the test accuracy of the process technology feature prediction model; Compare the accuracy value corresponding to the test feature data with the preset accuracy threshold, If the accuracy value corresponding to the test feature data is greater than or equal to the preset accuracy threshold, the preset process feature prediction model is obtained, wherein the preset process feature prediction model is the process feature prediction model; If the accuracy value corresponding to the test feature data is less than the preset accuracy threshold, the process feature prediction model is iteratively trained until the accuracy value corresponding to the test feature data is greater than or equal to the preset accuracy threshold, so as to obtain the preset process feature prediction model.

7. A data processing device for process feature recognition, characterized in that: include: A data acquisition module, used for acquiring data to be processed, wherein the data to be processed is data used to represent a workpiece to be processed, and the workpiece to be processed is a workpiece whose process features are to be identified; A feature extraction module, used for performing extraction processing based on three-dimensional features on the data to be processed to obtain three-dimensional feature data to be processed, wherein the three-dimensional feature data to be processed is data used to represent the three-dimensional object features of the workpiece to be processed; A process feature prediction module is used to perform process feature prediction processing on the three-dimensional feature data to be processed based on a preset process feature prediction model to obtain target process feature data, wherein the preset process feature prediction model is a graph neural network model, and the target process feature data is data used to represent the process features predicted for the workpiece to be processed.

8. The data processing device according to claim 7, characterized in that: The feature extraction module includes: A mapping module, used for performing mapping processing on the data to be processed based on the geometric elements of the object to obtain pre-processed data, wherein the pre-processed data is data used to represent the mapped geometric elements of the workpiece of the data to be processed; A first feature extraction module, used for performing feature extraction processing based on a first geometric element on the preprocessed data to obtain first geometric element feature data, wherein the first geometric element feature data is surface feature data for representing a workpiece to be processed; A second feature extraction module, used for performing feature extraction processing based on a second geometric element on the preprocessed data to obtain second geometric element feature data, wherein the second geometric element feature data is edge feature data for representing the workpiece to be processed; A graph construction module, used to perform graph construction processing on the preprocessed data to obtain constructed graph data; A three-dimensional feature extraction module is used to obtain the three-dimensional feature data to be processed according to the first geometric element feature data, the second geometric element feature data and the construction image data.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable the computer to execute the data processing method for process feature identification according to any one of claims 1 to 6.

10. An electronic device, characterized in that: include: at least one processor; And a memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor executes the data processing method for process feature identification described in any one of claims 1-6.

Images