Three-dimensional curve frame generation method and device, computer equipment and readable storage medium

Abstract

The invention relates to a three-dimensional curvilinear frame generation method and device, computer equipment and a readable storage medium. The method comprises the following steps: acquiring initial image data; performing feature coding on the initial image data to obtain initial image features corresponding to the initial image data; based on the initial image features, carrying out distribution mapping on the sampled noise data to obtain a three-dimensional curve frame potential representation corresponding to the initial image data; performing feature decoding on the three-dimensional curve frame potential representation to obtain a three-dimensional curve frame structure and curve potential representation corresponding to the initial image data; performing feature decoding on the curve potential representation to obtain a three-dimensional curve frame curve corresponding to the initial image data; and based on the three-dimensional curvilinear frame structure and the three-dimensional curvilinear frame curve, generating a target three-dimensional curvilinear frame corresponding to the initial image data. By adopting the method, the generation accuracy of the three-dimensional curve frame can be improved.

Description

Technical Field

[0001] The present application relates to the field of model generation technology, and in particular to a three-dimensional curve frame generation method, device, computer equipment, computer-readable storage medium and computer program product. Background Art

[0002] In the field of computer-aided design, 3D wireframe is an important form of expression for preliminary design, which has an intuitive display effect on the structure and layout of objects. 3D wireframe is a 3D shape representation method, which uses a network structure composed of continuous lines and discrete topological connections to express the geometric form of 3D objects.

[0003] When generating a 3D wireframe, the prior art usually generates a corresponding 3D wireframe based on a 3D object model in point cloud or image data based on a deep learning method. However, when processing sparse or incomplete input data, the prior art methods are prone to generating incomplete or structurally incorrect 3D wireframes. Furthermore, the prior art methods are also limited to regular 3D object models or simple geometric shapes. For 3D object models with complex curves, the generated 3D curve frames have large errors, resulting in the problem of low generation accuracy of 3D curve frames. Summary of the invention

[0004] Based on this, it is necessary to provide a three-dimensional curve frame generation method, device, computer equipment, computer-readable storage medium and computer program product that can improve the generation accuracy of the three-dimensional curve frame in response to the above technical problems.

[0005] In a first aspect, the present application provides a method for generating a three-dimensional curve frame, comprising:

[0006] Obtaining initial image data;

[0007] Performing feature encoding on the initial image data to obtain initial image features corresponding to the initial image data;

[0008] Based on the initial image features, the sampled noise data is distributed mapped to obtain the three-dimensional curve box potential representation corresponding to the initial image data;

[0009] Decoding the three-dimensional curve box potential representation to obtain the three-dimensional curve box structure and curve potential representation corresponding to the initial image data;

[0010] Decoding the curve potential representation to obtain a three-dimensional curved frame curve corresponding to the initial image data;

[0011] Based on the three-dimensional curve frame structure and the three-dimensional curve frame curve, a target three-dimensional curve frame corresponding to the initial image data is generated.

[0012] In a second aspect, the present application also provides a three-dimensional curve frame generation device, comprising:

[0013] An acquisition module, used for acquiring initial image data;

[0014] An encoding module, used for performing feature encoding on the initial image data to obtain initial image features corresponding to the initial image data;

[0015] A distribution mapping module is used to perform distribution mapping on the sampled noise data based on the initial image features to obtain a three-dimensional curve box potential representation corresponding to the initial image data;

[0016] A structure decoding module is used to perform feature decoding on the three-dimensional curve box potential representation to obtain the three-dimensional curve box structure and curve potential representation corresponding to the initial image data;

[0017] A curve decoding module is used to decode the characteristics of the curve potential representation to obtain a three-dimensional curve box curve corresponding to the initial image data;

[0018] The wireframe generation module is used to generate a target three-dimensional curve frame corresponding to the initial image data based on the three-dimensional curve frame structure and the three-dimensional curve frame curve.

[0019] In a third aspect, the present application further provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0020] Obtaining initial image data;

[0021] Performing feature encoding on the initial image data to obtain initial image features corresponding to the initial image data;

[0022] Based on the initial image features, the sampled noise data is distributed mapped to obtain the three-dimensional curve box potential representation corresponding to the initial image data;

[0023] Decoding the three-dimensional curve box potential representation to obtain the three-dimensional curve box structure and curve potential representation corresponding to the initial image data;

[0024] Decoding the curve potential representation to obtain a three-dimensional curved frame curve corresponding to the initial image data;

[0025] Based on the three-dimensional curve frame structure and the three-dimensional curve frame curve, a target three-dimensional curve frame corresponding to the initial image data is generated.

[0026] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the following steps are implemented:

[0027] Obtaining initial image data;

[0028] Performing feature encoding on the initial image data to obtain initial image features corresponding to the initial image data;

[0029] Based on the initial image features, the sampled noise data is distributed mapped to obtain the three-dimensional curve box potential representation corresponding to the initial image data;

[0030] Decoding the three-dimensional curve box potential representation to obtain the three-dimensional curve box structure and curve potential representation corresponding to the initial image data;

[0031] Decoding the curve potential representation to obtain a three-dimensional curved frame curve corresponding to the initial image data;

[0032] Based on the three-dimensional curve frame structure and the three-dimensional curve frame curve, a target three-dimensional curve frame corresponding to the initial image data is generated.

[0033] In a fifth aspect, the present application further provides a computer program product, including a computer program, which implements the following steps when executed by a processor:

[0034] Obtaining initial image data;

[0035] Performing feature encoding on the initial image data to obtain initial image features corresponding to the initial image data;

[0036] Based on the initial image features, the sampled noise data is distributed mapped to obtain the three-dimensional curve box potential representation corresponding to the initial image data;

[0037] Decoding the three-dimensional curve box potential representation to obtain the three-dimensional curve box structure and curve potential representation corresponding to the initial image data;

[0038] Decoding the curve potential representation to obtain a three-dimensional curved frame curve corresponding to the initial image data;

[0039] Based on the three-dimensional curve frame structure and the three-dimensional curve frame curve, a target three-dimensional curve frame corresponding to the initial image data is generated.

[0040] The above-mentioned three-dimensional curve frame generation method, device, computer equipment, computer-readable storage medium and computer program product obtain the initial image features by feature encoding the initial image data, and based on the initial image features, distribute and map the sampled noise data to obtain the three-dimensional curve frame potential representation corresponding to the initial image data, so that the three-dimensional curve frame potential representation can integrate the curve frame structure and curve feature information about the three-dimensional curve frame in the initial image data, thereby ensuring the data integrity of the three-dimensional curve frame represented by the three-dimensional curve frame potential representation; since the construction of the three-dimensional curve frame requires the dependency relationship between the curve frame structure and the curve of the curve frame, the reconstructed initial image data corresponding to the three-dimensional curve frame is synchronously obtained by feature decoding the three-dimensional curve frame potential representation. The three-dimensional curve frame structure and curve potential representation avoid the separate processing of encoding and decoding reconstruction of the three-dimensional curve frame structure based on the initial image data and encoding and decoding reconstruction of the curve frame curve; compared with the separate processing of the initial image data in the curve frame structure and the curve frame curve (encoding and decoding reconstruction), the initial image features are obtained by encoding the initial image data, and then the three-dimensional curve frame potential representation corresponding to the initial image features is synchronously decoded to obtain the reconstructed three-dimensional curve frame structure and curve potential representation, which ensures the accuracy of the three-dimensional curve frame structure and curve potential representation, and then when the curve potential representation is decoded and reconstructed, the accuracy of the three-dimensional curve frame curve is guaranteed, thereby improving the generation accuracy of the three-dimensional curve frame. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the drawings required for use in the embodiments of the present application or related technical descriptions will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0042] Figure 1 A diagram of an application environment of a three-dimensional curve frame generation method in an embodiment;

[0043] Figure 2 A schematic diagram of a flow chart of a method for generating a three-dimensional curve frame in one embodiment;

[0044] Figure 3 A schematic diagram of training a wireframe processing model in one embodiment;

[0045] Figure 4 A schematic diagram of a standard curve frame curve in one embodiment;

[0046] Figure 5 A training diagram of a curve processing model in one embodiment;

[0047] Figure 6A schematic diagram of training a distribution mapping model in one embodiment;

[0048] Figure 7 A schematic diagram of a three-dimensional curve frame generation model in one embodiment;

[0049] Figure 8 is a structural block diagram of a three-dimensional curve frame generating device in one embodiment;

[0050] Fig. 9 is an internal structure diagram of a computer device in one embodiment;

[0051] Fig.10 FIG. 4 is a diagram showing the internal structure of a computer device in another embodiment. DETAILED DESCRIPTION

[0052] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0053] The three-dimensional curve frame generation method provided in the embodiment of the present application can be applied to Figure 1 In the application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be integrated on the server 104, or it can be placed on the cloud or other network servers. The server 104 can obtain the initial image data through the terminal 102; the server 104 performs feature encoding on the initial image data to obtain the initial image features corresponding to the initial image data; the server 104 performs distribution mapping on the sampled noise data based on the initial image features to obtain the three-dimensional curve box potential representation corresponding to the initial image data; the server 104 performs feature decoding on the three-dimensional curve box potential representation to obtain the three-dimensional curve box structure and curve potential representation corresponding to the initial image data; the server 104 performs feature decoding on the curve potential representation to obtain the three-dimensional curve box curve corresponding to the initial image data; the server 104 generates the target three-dimensional curve box corresponding to the initial image data based on the three-dimensional curve box structure and the three-dimensional curve box curve. Among them, the terminal 102 can be, but is not limited to, various personal computers, laptops, smart phones, tablet computers, etc. The server 104 can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server providing cloud computing services.

[0054] In an exemplary embodiment, Figure 2 As shown, a three-dimensional curve frame generation method is provided, and the method is applied to Figure 1 Taking the server 104 in the example as an example, the following steps are included:

[0055] Step 202: Acquire initial image data.

[0056] Step 204 , feature encoding is performed on the initial image data to obtain initial image features corresponding to the initial image data.

[0057] The initial image data refers to data with a three-dimensional image, such as point cloud data, image data, etc. The point cloud data can be sparse point cloud data or incomplete point cloud data, and the image data can be a sketch, a drawing, etc. The initial image feature refers to a feature vector of a specific dimension obtained after encoding the initial image data.

[0058] Exemplarily, the server obtains initial image data in response to a request for generating a three-dimensional curve frame sent by the terminal. The initial image data may include one or more data having a three-dimensional image, such as point cloud data and image data. A three-dimensional image is, for example, a curve frame image representing a three-dimensional object. The server encodes the initial image data by calling a corresponding image encoder according to the data type of the initial image data (such as point cloud data and image data), performing feature encoding on the three-dimensional image in the initial image data, and obtaining initial image features of a specific dimension. Generally, the initial image features are feature vectors of 1024 dimensions.

[0059] Step 206 , based on the initial image features, distribution mapping is performed on the sampled noise data to obtain a three-dimensional curve box potential representation corresponding to the initial image data.

[0060] The sampling noise data may be random noise that obeys a normal distribution, such as Gaussian noise, for providing data sampling. Distribution mapping refers to the process of mapping a distribution (such as a simple distribution: a standard normal distribution) to a target data distribution (such as the distribution of complex data such as images and text). The three-dimensional curve box potential representation is identification data representing the three-dimensional curve box, which is used as data representing the three-dimensional curve box for feature decoding, representing the wireframe information of the three-dimensional curve box, including information such as the curve box structure and the curve box curve.

[0061] Exemplarily, the server obtains noise data of a preset dimension, generally Gaussian noise. The preset dimension corresponding to the obtained noise data is consistent with the feature dimension corresponding to the initial image feature, for example, 1024 dimensions. Then the server uses the three-dimensional curve box shape corresponding to the initial image feature as a constraint condition, maps the normal distribution corresponding to the sampled noise data to the target data distribution that satisfies the constraint condition, samples the target data distribution, and obtains the three-dimensional curve box potential representation corresponding to the initial image data.

[0062] Step 208: feature decoding is performed on the three-dimensional curve frame potential representation to obtain the three-dimensional curve frame structure and curve potential representation corresponding to the initial image data.

[0063] The three-dimensional curve frame structure refers to information describing the connection relationship of the curve frame curve, such as the topological structure corresponding to the initial image data. The curve potential representation is data representing the curve frame curve for feature decoding, which can be a feature representing the geometric shape of the curve frame curve, such as a wavy geometric shape, an arc-shaped geometric shape, etc.

[0064] Exemplarily, the server calls a pre-trained target wireframe processing model, and the target wireframe processing model is used to decode and reconstruct the input three-dimensional curve frame potential representation into a three-dimensional curve frame structure and a curve potential representation. The decoding network (decoder) in the target wireframe processing model may be pre-trained, and the server inputs the three-dimensional curve frame potential representation into the trained decoding network in the target wireframe processing model, and performs feature decoding on the three-dimensional curve frame potential representation through the decoding network to obtain the three-dimensional curve frame structure and curve potential representation corresponding to the initial image. The target wireframe processing model also includes an encoder, which can encode the input three-dimensional curve frame structure and curve potential representation into a three-dimensional curve frame potential representation. The encoder is also used to train the decoder through the three-dimensional curve frame potential representation encoded by the encoder during the training process of the target wireframe processing model, so that the decoder decodes the three-dimensional curve frame potential representation to obtain an accurate three-dimensional curve frame structure and curve potential representation.

[0065] In an exemplary embodiment, the dimension of the potential representation of the three-dimensional curve box can be adjusted according to different input data, such as adjusting according to the data volume of the input data or adjusting according to the data type of the input data.

[0066] Step 210 , feature decoding is performed on the curve potential representation to obtain a three-dimensional curve frame curve corresponding to the initial image data.

[0067] The three-dimensional curve frame curve refers to a curve line with a geometric shape. The curve potential representation can be decoded to obtain a three-dimensional curve frame curve.

[0068] Exemplarily, after obtaining the potential representation of the curve, the server calls a pre-trained target curve processing model. The target curve processing model is used to decode the input potential representation of the curve into a three-dimensional curve frame curve. The decoding network in the target curve decoding model may be pre-trained. The server then inputs the curve potential representation into the decoding network (decoder) in the target curve processing model, and decodes the curve potential representation into a corresponding three-dimensional curve frame curve through the decoding network. The three-dimensional curve frame curve includes multiple and various types, and a straight line is also a type of curve. The target curve processing model also includes an encoder, which can encode the input three-dimensional curve frame curve into a curve potential representation. The encoder is also used to train the decoder through the curve potential representation encoded by the encoder during the training process of the target curve processing model, so that the decoder decodes the curve potential representation to obtain an accurate three-dimensional curve frame curve. The target curve decoding model can be a network model integrated with the target image processing model, or it can be an independent network model.

[0069] Step 212: Generate a target three-dimensional curve frame corresponding to the initial image data based on the three-dimensional curve frame structure and the three-dimensional curve frame curve.

[0070] Exemplarily, the three-dimensional curve frame structure includes the endpoint coordinates and adjacent relationships of the three-dimensional curve frame curves. The server determines adjacent three-dimensional curve frame curves based on the adjacent relationships, and connects the adjacent three-dimensional curve frame curves according to the endpoint coordinates to obtain the target three-dimensional curve frame corresponding to the initial image data. For example, according to the adjacent relationship, it is determined that the endpoint A1 of curve A is adjacent to the endpoint B1 of curve B and the endpoint C1 of curve C, then curves A, B, and C are connected according to the endpoint coordinates of the endpoints A1 and A2 of curve A, the endpoints B1 and B2 of curve B, and the endpoints C1 and C2 of curve C.

[0071] In the above three-dimensional curve frame generation method, the initial image features are obtained by feature encoding the initial image data, and based on the initial image features, the sampled noise data is distributed and mapped to obtain the three-dimensional curve frame potential representation corresponding to the initial image data, so that the three-dimensional curve frame potential representation can integrate the feature information of the curve frame structure and the curve of the three-dimensional curve frame in the initial image data, thereby ensuring the data integrity of the three-dimensional curve frame represented by the three-dimensional curve frame potential representation; since the construction of the three-dimensional curve frame requires the dependency relationship between the curve frame structure and the curve of the curve frame, the three-dimensional curve frame structure and the curve potential representation corresponding to the reconstructed initial image data are synchronously obtained by feature decoding the three-dimensional curve frame potential representation. Representation, avoiding the separate processing of encoding and decoding reconstruction of the three-dimensional curve frame structure based on the initial image data and encoding and decoding reconstruction of the curve frame curve; compared with the separate processing of the initial image data in the curve frame structure and the curve frame curve (encoding and decoding reconstruction), the initial image features are obtained by encoding the initial image data, and then the three-dimensional curve frame potential representation corresponding to the initial image features is synchronously decoded to obtain the reconstructed three-dimensional curve frame structure and curve potential representation, which ensures the accuracy of the three-dimensional curve frame structure and curve potential representation, and then when the curve potential representation is decoded and reconstructed, the accuracy of the three-dimensional curve frame curve is guaranteed, thereby improving the generation accuracy of the three-dimensional curve frame.

[0072] In an exemplary embodiment, the 3D curve frame generating method further includes:

[0073] Obtaining a first training curve frame structure and a first training curve frame curve corresponding to the first training three-dimensional curve frame;

[0074] Performing feature encoding on the first training curve frame structure to obtain first training structure features, and performing feature encoding on the first training curve frame curve to obtain a first training curve potential representation;

[0075] Fusing the first training structural feature and the first training curve potential representation to obtain a first training curve box potential representation;

[0076] Inputting the first training curve frame potential representation into the initial wireframe processing model to obtain a first prediction curve frame structure and a first prediction curve potential representation corresponding to the first training three-dimensional curve frame;

[0077] Based on the first training curve frame structure, the first training curve potential representation, the first prediction curve frame structure, and the first prediction curve potential representation, adjusting model parameters of the initial wireframe processing model until a convergence condition is met to obtain a target wireframe processing model;

[0078] Step 208, feature decoding is performed on the three-dimensional curve frame potential representation to obtain the three-dimensional curve frame structure and curve potential representation corresponding to the initial image data, including:

[0079] The three-dimensional curve box potential representation is input into the target wireframe processing model to obtain the three-dimensional curve box structure and curve potential representation corresponding to the initial image data.

[0080] Among them, the first training three-dimensional curve frame is the training data for training the initial wireframe processing model, which is represented in the form of a three-dimensional curve frame. The first training curve frame structure refers to the topological structure corresponding to the first training three-dimensional curve frame. The first training curve frame curve refers to the curve line in the first training three-dimensional curve frame, including curve lines of different geometric shapes. The first training structure feature is a feature vector characterizing the first training curve frame structure. The first training curve potential representation is data characterizing the geometric shape of the first training curve frame curve. The first training curve frame potential representation refers to the three-dimensional curve frame potential representation used to train the initial wireframe processing model. The first predicted curve frame structure refers to the curve frame structure predicted (decoded) by the initial wireframe processing model according to the first training curve frame potential representation. The first predicted curve potential representation is the curve potential representation predicted (decoded) by the initial wireframe processing model according to the first training curve frame potential representation.

[0081] Exemplarily, an initial wireframe processing model to be trained is pre-deployed in the server, which is used to generate a three-dimensional curve frame structure and a curve potential representation. Before training the initial wireframe processing model, the server obtains the first training curve frame structure and the first training curve frame curve corresponding to the first training three-dimensional curve frame, and performs feature encoding on the first training curve frame curve to obtain the first training curve potential representation, which can be performed by using an encoder in the trained target curve processing model to perform feature encoding on the first training curve frame curve to obtain the first training curve potential representation. Then, feature extraction is performed on the first training curve frame structure, such as converting the first training curve frame structure into a data list, which can be based on each vertex in the topological network corresponding to the first training curve frame structure (i.e., the endpoints of the first training curve frame curve) and the adjacent vertices connected to each vertex (i.e., the endpoints of the adjacent first training curve frame curve in the first training three-dimensional curve frame), and the endpoint coordinate data is generated according to the endpoint coordinates of each endpoint of the first training curve frame curve (the three-dimensional coordinates of the endpoint), and the first training structure feature is obtained according to the data list and the endpoint coordinate data converted from the first training curve frame structure. In order to ensure that the topological structures of different wireframes have a uniform vertex arrangement distribution, a breadth-first search may be used to sort the adjacency list. The server fuses the first training structure feature and the first training curve potential representation, and may concatenate the first training structure feature and the first training curve potential representation to obtain a concatenated first training curve box potential representation. The first training curve box potential representation is used to train the initial wireframe processing model.

[0082] The server inputs the first training curve frame potential representation into the initial wireframe processing model to obtain the first prediction curve frame structure and the first prediction curve potential representation corresponding to the first training three-dimensional curve frame. Then, the server iteratively adjusts the model parameters of the initial wireframe processing model according to the loss between the first training curve frame structure and the first prediction curve frame structure, and the loss between the first training curve potential representation and the first prediction curve potential representation, until the convergence condition is met, and obtains the trained target wireframe processing model. The initial wireframe processing model and the target wireframe processing model can be variational autoencoder models.

[0083] In the process of generating the target three-dimensional curve frame corresponding to the initial image data, after obtaining the potential representation of the three-dimensional curve frame corresponding to the initial image data, the server inputs the potential representation of the three-dimensional curve frame into the target wireframe processing model to obtain the three-dimensional curve frame structure and curve potential representation corresponding to the initial image data.

[0084] In this embodiment, the initial wireframe processing model is trained by using the first training curve frame potential representation after the first training structure feature and the first training curve potential representation to achieve synchronous decoding of the three-dimensional curve frame structure and the curve potential representation by the target wireframe processing model, thereby ensuring the accuracy of the three-dimensional curve frame structure and the curve potential representation. In addition, the three-dimensional curve frame structure (topological structure) is converted into an adjacency list, which simplifies the representation of the topological structure and reduces redundant information, thereby reducing the amount of calculation of the wireframe processing model and improving the processing efficiency of the wireframe processing model.

[0085] In an exemplary embodiment, the initial wireframe processing model includes an encoder and a decoder; inputting the first training curve frame potential representation into the initial wireframe processing model to obtain a first prediction curve frame structure and a first prediction curve potential representation corresponding to the first training three-dimensional curve frame includes:

[0086] Inputting the first training curve box potential representation into an encoder in the initial wireframe processing model to obtain an intermediate training curve box potential representation;

[0087] Inputting the intermediate training curve frame potential representation into a decoder in the initial wireframe processing model to obtain a first prediction curve frame structure and a first prediction curve potential representation corresponding to the first training three-dimensional curve frame;

[0088] The three-dimensional curve frame potential representation is input into the target wireframe processing model to obtain the three-dimensional curve frame structure and curve potential representation corresponding to the initial image data, including:

[0089] The three-dimensional curve box potential representation is input into the decoder in the target wireframe processing model to obtain the three-dimensional curve box structure and curve potential representation corresponding to the initial image data.

[0090] Exemplarily, the initial wireframe processing model includes an encoder and a decoder. In the training process of the initial wireframe processing model, the encoder and the decoder in the initial wireframe processing model are specifically trained. The server inputs the first training curve frame potential representation into the encoder in the initial wireframe processing model to obtain an intermediate training curve frame potential representation. The intermediate training curve frame potential representation is a wireframe potential representation of a fixed length obtained after encoding by the encoder, such as a fixed length of 64x16. The encoder in the initial wireframe processing model can be understood as being used to encode the input wireframe potential representation in a fixed length dimension. Then the server inputs the intermediate training curve frame potential representation into the decoder in the initial wireframe processing model for decoding, and obtains a first prediction curve frame structure and a first prediction curve potential representation corresponding to the first training three-dimensional curve frame.

[0091] The server optimizes the encoder in the initial wireframe processing model according to the loss between the first training curve frame structure and the first prediction curve frame structure, and optimizes the decoder in the initial wireframe processing model according to the loss between the first training curve potential representation and the first prediction curve potential representation, until the convergence condition is met to obtain the target wireframe processing model.

[0092] In the process of generating the target three-dimensional curve frame corresponding to the initial image data, after obtaining the potential representation of the three-dimensional curve frame corresponding to the initial image data, the server inputs the potential representation of the three-dimensional curve frame into the decoder in the target wireframe processing model for decoding to obtain the three-dimensional curve frame structure and curve potential representation corresponding to the initial image data.

[0093] In an exemplary embodiment, Figure 3 As shown, a training schematic diagram of a wireframe processing model is provided. Figure 3 The input in is the first training three-dimensional curve frame, and the arrow points to the first training curve frame structure (i.e., the topological network in the figure) and the first training curve frame curve corresponding to the first training three-dimensional curve frame. The first training curve frame curve is encoded by the trained target curve processing model to obtain the potential representation of the first training curve, such as Figure 3 The Nx12 rectangular blocks in the figure represent the curve box potential representation of N wireframe curves; the first training curve box structure is converted into an adjacency list (such as Figure 3 Nx48 rectangular blocks) and endpoint coordinate data (such as Figure 3 Then, the first training structure feature and the first training curve potential representation are fused to obtain the first training curve box potential representation, as shown in Figure 3 The NxCw rectangular block Zcurve in the image is input into the encoder of the wireframe processing model, and the first training curve box potential representation is input into the encoder of the wireframe processing model, such as Figure 3The Cross Attention model in , where the 64xCw rectangular block represents the input vector Query of the cross attention model, and the intermediate training curve box potential representation with a fixed length of 64x16 is obtained, as shown in Figure 3 The intermediate training curve box potential representation is then input into the decoder in the initial wireframe processing model, as Figure 3 The cross attention model (Cross Attention) and self attention model (self Attention) pointed to by the potential representation of the intermediate training curve box in the MxCw, wherein the rectangular block of MxCw represents the input vector Query of the cross attention model, outputs the prediction result, and obtains the first prediction curve box structure (including the predicted adjacency list and endpoint coordinate data) and the first prediction curve potential representation corresponding to the first training three-dimensional curve box. According to the loss between the first training curve box structure and the first prediction curve box structure, the encoder in the initial wireframe processing model is optimized, and according to the loss between the first training curve potential representation and the first prediction curve potential representation, the decoder in the initial wireframe processing model is optimized, which can be to calculate the cross entropy loss between the first training curve box structure and the first prediction curve box structure to optimize the encoder, and calculate the KL divergence loss between the first training curve potential representation and the first prediction curve potential representation to constrain the potential representation to optimize the decoder until the convergence condition is met to obtain the target wireframe processing model.

[0094] In this embodiment, the curves are encoded through the cross-attention mechanism, which can handle more complex and diverse curve geometries and ensure the accuracy of the generation of the three-dimensional curve box.

[0095] In an exemplary embodiment, the 3D curve frame generating method further includes:

[0096] Obtaining a third training curve box curve and a third training curve potential representation;

[0097] Inputting the third training curve frame curve into an encoder in the initial curve processing model to obtain a third prediction curve latent representation;

[0098] Inputting the third prediction curve potential representation into a decoder in the initial curve processing model to obtain a third prediction curve frame curve;

[0099] Based on the third training curve potential representation, the third prediction curve potential representation, the third training curve frame curve, and the third prediction curve frame curve, adjusting the model parameters of the initial curve processing model until a convergence condition is met to obtain a target curve processing model;

[0100] Step 210, feature decoding is performed on the curve potential representation to obtain a three-dimensional curve frame curve corresponding to the initial image data, including:

[0101] The curve potential representation is input into the decoder in the target curve processing model to obtain the three-dimensional curved box curve corresponding to the initial image data.

[0102] The third training curve frame curve is a wireframe curve used to train the initial curve processing model. The third training curve potential representation is a real curve potential representation obtained by pre-encoding the third training curve frame curve. The third prediction curve potential representation is a curve potential representation obtained by encoding the third training curve frame curve by the encoder of the initial curve processing model. The third prediction curve frame curve is a wireframe curve obtained by decoding the third prediction curve potential representation by the decoder of the initial curve processing model.

[0103] Exemplarily, an initial curve processing model to be trained is pre-deployed in the server, and the curve processing model can encode the input three-dimensional curve frame curve into a curve potential representation, and decode the encoded curve potential representation and reconstruct the three-dimensional curve frame curve. The initial curve processing model includes an encoder and a decoder. In the training process of the initial curve processing model, the encoder and the decoder in the initial curve processing model are specifically trained.

[0104] During the training of the initial curve processing model, the server obtains the third training curve frame curve and the third training curve potential representation, inputs the third training curve frame curve into the encoder in the initial curve processing model for feature encoding, and obtains the third prediction curve potential representation. Then, the third prediction curve potential representation is input into the decoder in the initial curve processing model to obtain the third prediction curve frame curve.

[0105] The server iteratively optimizes the encoder in the initial curve processing model according to the loss between the potential representation of the third training curve and the potential representation of the third prediction curve, and iteratively optimizes the decoder in the initial curve processing model according to the loss between the third training curve frame curve and the third prediction curve frame curve until the convergence condition is met to obtain the target curve processing model. The encoder and decoder in the initial curve processing model may be optimized using mean square error loss. The initial curve processing model and the target curve processing model may be variational autoencoder models.

[0106] In the process of generating the target three-dimensional curve frame corresponding to the initial image data, after obtaining the curve potential representation corresponding to the three-dimensional curve frame potential representation, the server inputs the curve potential representation into the decoder in the target curve processing model for decoding to obtain the three-dimensional curve frame curve corresponding to the initial image data.

[0107] In this embodiment, by training the encoder and decoder in the initial curve processing model, the accuracy of the potential representation of the curve obtained by encoding and the accuracy of the curve frame curve obtained by decoding are guaranteed.

[0108] In an exemplary embodiment, obtaining a third training curve frame curve includes:

[0109] Get the initial curve frame curve;

[0110] Based on the preset starting point and the preset end point, the starting point and the end point of each curve in the initial curve frame curve are unified to obtain a standard curve frame curve;

[0111] The standard curve frame curve is discretized to obtain a third training curve frame curve.

[0112] Exemplarily, the server obtains an initial curve frame curve, and performs geometric transformations on each curve in the initial curve frame curve according to the starting coordinates of a preset starting point and the end coordinates of a preset end point, such as translation, rotation, and scaling, so that the starting point (i.e., the starting end point) of each curve coincides with the preset starting point, and the end point (i.e., the end end point) of each curve coincides with the preset end point, and each curve with a unified starting position and end point position is obtained, and each curve with a unified starting position and end point position is used as a standard curve frame curve. The schematic diagram of the standard curve frame curve is shown as follows: Figure 4 As shown, Figure 4 The left image in the middle shows the curves before the positions are unified, and the right image shows the curves after the starting and ending positions are unified. The starting and ending points of each curve are fixed at the same position (the preset starting point coordinates [-1, 0, 0] and the preset ending point coordinates [1, 0,0]).

[0113] Then the server divides each standard curve frame curve into 256 discrete sampling points to obtain a third training curve frame curve.

[0114] In this embodiment, by adopting standardization processing, the starting point and the end point are fixed at the same position, thereby eliminating redundant information caused by different scales and directions of different curves, thereby enhancing the training stability of the initial curve processing model.

[0115] In an exemplary embodiment, Figure 5 As shown, a training schematic diagram of a curve processing model is provided. Figure 5 The input curve is the third training curve frame curve after position unification and discretization. The third training curve frame curve is converted into a 256-characteristic vector, such as Figure 5 Then, the feature vector of the third training curve frame curve is downsampled, such as Figure 5 The 64xC rectangular block in is used as the input vector Query of the encoder (Cross Attention model) in the initial curve processing model. Figure 5K and V in it represent the input vector Key and input vector Value of the encoder (Cross Attention model).

[0116] The feature vector of the third training curve frame curve is input into the encoder in the initial curve processing model (such as Figure 5 The Cross Attention model on the left in the figure is used to encode the features of the third training curve box curve through the encoder to obtain the potential representation of the third prediction curve, and then use a one-dimensional convolution layer (such as Figure 5 The residual network in ResNet1D) compresses the dimension of the potential representation of the third prediction curve to obtain the final potential representation of the third prediction curve (such as Figure 5 The obtained third prediction curve potential representation is input into the decoder in the initial curve processing model (such as Figure 5 The server iteratively optimizes the encoder and decoder in the initial curve processing model according to the third training curve potential representation, the third prediction curve potential representation, the third training curve frame curve, and the third prediction curve frame curve to obtain the target curve processing model.

[0117] In an exemplary embodiment, the 3D curve frame generating method further includes:

[0118] Obtaining a second training structural feature and a second training curve potential representation corresponding to the second training three-dimensional curve frame;

[0119] Fusing the second training structure feature and the second training curve potential representation to obtain a second training curve box potential representation corresponding to the second training three-dimensional curve box;

[0120] Inputting the second training curve box potential representation and the sampled noise data into the initial distribution mapping model to obtain the prediction curve box potential representation corresponding to the second training three-dimensional curve box;

[0121] Based on the second training curve box potential representation and the prediction curve box potential representation, adjusting the model parameters of the initial distribution mapping model until a convergence condition is met, thereby obtaining a target distribution mapping model;

[0122] Step 206, based on the initial wireframe potential representation, distribution mapping is performed on the sampled noise data to obtain a three-dimensional curve frame potential representation corresponding to the initial image data, including:

[0123] The initial wireframe latent representation and the sampled noise data are input into the target distribution mapping model to obtain the 3D curve box latent representation corresponding to the initial image data.

[0124] The second training three-dimensional curve frame is used to provide the second training structural features and the second training curve potential representation to train the input data of the initial distribution mapping model, which is represented in the form of a three-dimensional curve frame. The second training structural features are structural features obtained according to the curve frame structure corresponding to the second training three-dimensional curve frame. The second training curve potential representation is a curve potential representation obtained according to the curve frame curve encoding corresponding to the second training three-dimensional curve frame.

[0125] Exemplarily, an initial distribution mapping model to be trained is also deployed in the server. Before training the initial distribution mapping model, the server obtains the second training structural features and the second training curve potential representation corresponding to the second training three-dimensional curve frame. The second training curve potential representation can be obtained by encoding the wireframe curve of the second training three-dimensional curve frame by the encoder of the target curve processing model. The second training structural features and the second training curve potential representation are fused to obtain the second training curve frame potential representation corresponding to the second training three-dimensional curve frame. The second training curve frame potential representation can be obtained by encoding the fused potential representation according to a fixed length by the encoder of the target wireframe processing model after fusing the second training structural features and the second training curve potential representation. Then, the second training curve frame potential representation and the sampled noise data are input into the initial distribution mapping model to obtain the predicted curve frame potential representation corresponding to the second training three-dimensional curve frame. According to the loss between the second training curve frame potential representation and the predicted curve frame potential representation, the model parameters of the initial distribution mapping model are adjusted until the convergence condition is met to obtain the target distribution mapping model. Specifically, the target distribution mapping model may refer to a flow matching model, which includes a velocity field represented by a neural network, which is used to describe the instantaneous velocity of the potential representation, and converts Gaussian noise (sampled noise data) into a target data distribution by solving ordinary differential equations. In the optimization process of the flow matching model, the flow matching model can be optimized by minimizing the error of the velocity field to obtain an optimized flow matching model, namely, the target distribution mapping model.

[0126] In the process of generating the target three-dimensional curve frame corresponding to the initial image data, after obtaining the initial wireframe potential representation, the server inputs the initial wireframe potential representation into the distribution mapping model, and outputs the three-dimensional curve frame potential representation corresponding to the initial image data through the distribution mapping model. In an exemplary embodiment, Figure 6 As shown, a training schematic diagram of a distribution mapping model is provided. Figure 6The input in is the second training three-dimensional curve frame, and the second training curve frame structure and the second training curve frame curve corresponding to the second training three-dimensional curve frame are obtained. The second training curve frame structure is converted into a second training structure feature, and the second training curve frame curve is input into the encoder in the target curve processing model for encoding to obtain the second training curve potential representation. The second training structure feature and the second training curve potential representation are fused, and the fused potential representation is input into the encoder in the target wireframe processing model for encoding according to a fixed length to obtain the second training curve frame potential representation corresponding to the second training three-dimensional curve frame (such as Figure 6 The 64x16 rectangular blocks in ). Then the second training curve box potential representation and sampled noise data (normal distribution with mean 0 and standard deviation 1) are used. , recorded as N (0, 1)) input the initial distribution mapping model, and obtain the predicted curve box potential representation corresponding to the second training three-dimensional curve box (such as Figure 6 The dotted rectangle in ), according to the loss LOSS between the initial potential representation of the second training line and the potential representation of the prediction curve box, the initial distribution mapping model is optimized, which can be the velocity field vt, to obtain the target distribution mapping model.

[0127] In an exemplary embodiment, the flow matching model supports unconditional and conditional generation. In unconditional generation, the flow matching model generates a 3D wireframe from Gaussian noise; in conditional generation, the input image data or point cloud data is processed by a pre-trained feature extractor (such as DINOv2 (an unsupervised learning method based on the ViT model) and PointNet++ (a neural network for classifying and segmenting irregular point cloud data) and used as a constraint input to control the flow matching model to generate a 3D wireframe that meets the constraint.

[0128] In this embodiment, through the application of the latent flow matching model, the generation process is made smoother and continuous, the errors in the discrete generation method are avoided, the high accuracy and diversity of the generation results are ensured, and it can support unconditional and conditional generation and has stronger adaptability.

[0129] In an exemplary embodiment, Figure 7 As shown, a schematic diagram of a 3D curve frame generation model is provided. The initial image data is taken as input, and the initial image data includes sparse point cloud data, incomplete point cloud data, sketch image, etc. According to the encoder corresponding to the initial image data, the initial image data is feature encoded to obtain the initial wireframe potential representation corresponding to the initial image data; the initial wireframe potential representation and Gaussian noise are input into the stream matching model (target distribution mapping model), and the 3D curve frame potential representation corresponding to the initial image data is output through the stream matching model, as shown in FIG. Figure 7The 64x16 rectangular block Zw in the middle; the 3D curve frame potential representation is input into the decoder in the target wireframe processing model for feature decoding, and the 3D curve frame structure corresponding to the initial image data is obtained (such as Figure 7 The topological structure of the mesh) and the curve potential representation (such as Figure 7 The Nx12 rectangular block Zcurve in the target curve processing model is input into the decoder of the curve potential representation for feature decoding to obtain the three-dimensional curve frame curve corresponding to the initial image data; then, according to the three-dimensional curve frame structure and the three-dimensional curve frame curve, the target three-dimensional curve frame corresponding to the initial image data is generated.

[0130] Among them, the target curve processing model can be a curve variational autoencoder, which is used to encode various types of geometric curves into curve potential representations during the training process, and can adapt to the geometric features of different curves, such as straight lines, circles, Bezier curves, etc. The target wireframe processing model can be a box variational autoencoder, which is used to combine the curve potential representation (curve geometric information) generated by the curve variational autoencoder with the topological information during the training process to generate a global fixed-length three-dimensional curve box potential representation, so as to reconstruct the complete three-dimensional wireframe through subsequent encoding. Flow matching maps Gaussian noise to the target data distribution (i.e., the distribution of the target wireframe potential representation) through a velocity field of a neural network based on the three-dimensional curve box potential representation generated by the wireframe variational autoencoder to output the three-dimensional curve box potential representation.

[0131] In the decoding process of the target curve processing model, the curve potential representation is input into the upsampling convolutional network, and it is converted into a continuous three-dimensional curve using the cross-attention mechanism and multi-layer perceptron. In the decoding process of the target wireframe processing model, a set of learnable query vectors (such as Figure 3 The rectangular block MxCw representing the input vector Query in the figure and the three-dimensional curved box potential representation (such as Figure 3 The 64x16 rectangular blocks in the figure perform cross-attention calculations to extract the features of each curve and generate the corresponding adjacency list, endpoint coordinates, and curve potential representation.

[0132] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0133] Based on the same inventive concept, the embodiment of the present application also provides a 3D curve frame generation device for implementing the 3D curve frame generation method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more 3D curve frame generation device embodiments provided below can refer to the limitations of the 3D curve frame generation method above, and will not be repeated here.

[0134] In an exemplary embodiment, Figure 8 As shown, a three-dimensional curve frame generation device 800 is provided, comprising: an acquisition module 802, an encoding module 804, a distribution mapping module 806, a structure decoding module 808, a curve decoding module 810 and a wireframe generation module 812, wherein:

[0135] An acquisition module 802 is used to acquire initial image data;

[0136] The encoding module 804 is used to perform feature encoding on the initial image data to obtain an initial wireframe potential representation corresponding to the initial image data;

[0137] A distribution mapping module 806 is used to perform distribution mapping on the sampled noise data based on the initial image features to obtain a three-dimensional curve box potential representation corresponding to the initial image data;

[0138] The structure decoding module 808 is used to perform feature decoding on the three-dimensional curve box potential representation to obtain the three-dimensional curve box structure and curve potential representation corresponding to the initial image data;

[0139] The curve decoding module 810 is used to perform feature decoding on the curve potential representation to obtain a three-dimensional curve frame curve corresponding to the initial image data;

[0140] The wireframe generation module 812 is used to generate a target three-dimensional curve frame corresponding to the initial image data based on the three-dimensional curve frame structure and the three-dimensional curve frame curve.

[0141] In an exemplary embodiment, the three-dimensional curve frame generation device 800 is further used to obtain a first training curve frame structure and a first training curve frame curve corresponding to the first training three-dimensional curve frame; feature encode the first training curve frame structure to obtain a first training structure feature, feature encode the first training curve frame curve to obtain a first training curve potential representation; fuse the first training structure feature and the first training curve potential representation to obtain the first training curve frame potential representation; input the first training curve frame potential representation into the initial wireframe processing model to obtain a first prediction curve frame structure and a first prediction curve potential representation corresponding to the first training three-dimensional curve frame; based on the first training curve frame structure, the first training curve potential representation, the first prediction curve frame structure and the first prediction curve potential representation, adjust the model parameters of the initial wireframe processing model until the convergence condition is met to obtain the target wireframe processing model; the structure decoding module 808 is also used to input the three-dimensional curve frame potential representation into the target wireframe processing model to obtain the three-dimensional curve frame structure and curve potential representation corresponding to the initial image data.

[0142] In an exemplary embodiment, the initial wireframe processing model includes an encoder and a decoder, and the three-dimensional curve frame generation device 800 is also used to input the first training curve frame potential representation into the encoder in the initial wireframe processing model to obtain the intermediate training curve frame potential representation; input the intermediate training curve frame potential representation into the decoder in the initial wireframe processing model to obtain the first predicted curve frame structure and the first predicted curve potential representation corresponding to the first training three-dimensional curve frame; the structure decoding module 808 is also used to input the three-dimensional curve frame potential representation into the decoder in the target wireframe processing model to obtain the three-dimensional curve frame structure and curve potential representation corresponding to the initial image data.

[0143] In an exemplary embodiment, the three-dimensional curve frame generation device 800 is also used to obtain a third training curve frame curve and a third training curve potential representation; input the third training curve frame curve into the encoder in the initial curve processing model to obtain a third prediction curve potential representation; input the third prediction curve potential representation into the decoder in the initial curve processing model to obtain a third prediction curve frame curve; based on the third training curve potential representation, the third prediction curve potential representation, the third training curve frame curve and the third prediction curve frame curve, adjust the model parameters of the initial curve processing model until the convergence condition is met to obtain the target curve processing model; the curve decoding module 810 is also used to input the curve potential representation into the decoder in the target curve processing model to obtain the three-dimensional curve frame curve corresponding to the initial image data.

[0144] In an exemplary embodiment, the acquisition module 802 is also used to obtain an initial curve frame curve; based on a preset starting point and a preset end point, the starting point and end point of each curve in the initial curve frame curve are unified to obtain a standard curve frame curve; the standard curve frame curve is discretized to obtain a third training curve frame curve.

[0145] In an exemplary embodiment, the three-dimensional curve frame generation device 800 is also used to obtain the second training structural features and the second training curve potential representation corresponding to the second training three-dimensional curve frame; fuse the second training structural features and the second training curve potential representation to obtain the second training curve frame potential representation corresponding to the second training three-dimensional curve frame; input the second training curve frame potential representation and the sampled noise data into the initial distribution mapping model to obtain the predicted curve frame potential representation corresponding to the second training three-dimensional curve frame; based on the second training curve frame potential representation and the predicted curve frame potential representation, adjust the model parameters of the initial distribution mapping model until the convergence condition is met to obtain the target distribution mapping model; the three-dimensional curve frame generation device 800 is also used to input the initial wireframe potential representation and the sampled noise data into the target distribution mapping model to obtain the three-dimensional curve frame potential representation corresponding to the initial image data.

[0146] Each module in the above three-dimensional curve frame generation device can be implemented in whole or in part by software, hardware and a combination thereof. Each module can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute the operations corresponding to each module.

[0147] In an exemplary embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as shown in FIG. Fig. 9 As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data such as three-dimensional wireframes. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a three-dimensional curve frame generation method is implemented.

[0148] In an exemplary embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as shown in FIG. Fig.10 As shown. The computer device includes a processor, a memory, an input / output interface, a communication interface, a display unit and an input device. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface, the display unit and the input device are connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be implemented through WIFI, a mobile cellular network, near field communication (NFC) or other technologies. When the computer program is executed by the processor, a three-dimensional curve frame generation method is implemented. The display unit of the computer device is used to form a visually visible picture, which can be a display screen, a projection device or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a button, trackball or touchpad set on the computer device shell, or an external keyboard, touchpad or mouse.

[0149] Those skilled in the art will understand that Figure 9-10 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0150] In an exemplary embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above-mentioned method embodiments when executing the computer program.

[0151] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0152] In one embodiment, a computer program product is provided, including a computer program, which implements the steps in the above method embodiments when executed by a processor.

[0153] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0154] A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, an artificial intelligence (AI) processor, etc., but are not limited to this.

[0155] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0156] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A three-dimensional curve frame generation method, characterized in that: The method comprises: Obtaining initial image data; Performing feature encoding on the initial image data to obtain initial image features corresponding to the initial image data; Based on the initial image features, distribution mapping is performed on the sampled noise data to obtain a three-dimensional curve box potential representation corresponding to the initial image data; Performing feature decoding on the three-dimensional curve frame potential representation to obtain a three-dimensional curve frame structure and a curve potential representation corresponding to the initial image data; Performing feature decoding on the curve potential representation to obtain a three-dimensional curve frame curve corresponding to the initial image data; Based on the three-dimensional curve frame structure and the three-dimensional curve frame curve, a target three-dimensional curve frame corresponding to the initial image data is generated.

2. The method according to claim 1, characterized in that The method further comprises: Obtaining a first training curve frame structure and a first training curve frame curve corresponding to the first training three-dimensional curve frame; Performing feature encoding on the first training curve frame structure to obtain a first training structure feature, and performing feature encoding on the first training curve frame curve to obtain a first training curve potential representation; fusing the first training structure feature and the first training curve potential representation to obtain a first training curve box potential representation; Inputting the first training curve frame potential representation into an initial wireframe processing model to obtain a first prediction curve frame structure and a first prediction curve potential representation corresponding to the first training three-dimensional curve frame; Based on the first training curve frame structure, the first training curve potential representation, the first prediction curve frame structure, and the first prediction curve potential representation, adjusting model parameters of the initial wireframe processing model until a convergence condition is met to obtain a target wireframe processing model; The feature decoding of the three-dimensional curve frame potential representation to obtain the three-dimensional curve frame structure and curve potential representation corresponding to the initial image data includes: The three-dimensional curve frame potential representation is input into the target wireframe processing model to obtain the three-dimensional curve frame structure and curve potential representation corresponding to the initial image data.

3. The method according to claim 2, characterized in that The initial wireframe processing model includes an encoder and a decoder; The step of inputting the first training curve frame potential representation into an initial wireframe processing model to obtain a first prediction curve frame structure and a first prediction curve potential representation corresponding to the first training three-dimensional curve frame includes: Inputting the first training curve box potential representation into an encoder in the initial wireframe processing model to obtain an intermediate training curve box potential representation; Inputting the intermediate training curve frame potential representation into a decoder in the initial wireframe processing model to obtain a first prediction curve frame structure and a first prediction curve potential representation corresponding to the first training three-dimensional curve frame; The step of inputting the three-dimensional curve frame potential representation into the target wireframe processing model to obtain the three-dimensional curve frame structure and curve potential representation corresponding to the initial image data includes: The three-dimensional curve frame potential representation is input into a decoder in the target wireframe processing model to obtain a three-dimensional curve frame structure and a curve potential representation corresponding to the initial image data.

4. The method according to claim 1, characterized in that: The method further comprises: Obtaining a third training curve box curve and a third training curve potential representation; Inputting the third training curve frame curve into an encoder in an initial curve processing model to obtain a third prediction curve potential representation; Inputting the third prediction curve potential representation into a decoder in the initial curve processing model to obtain a third prediction curve frame curve; Based on the third training curve potential representation, the third prediction curve potential representation, the third training curve frame curve and the third prediction curve frame curve, adjusting the model parameters of the initial curve processing model until a convergence condition is met to obtain a target curve processing model; The step of decoding the curve potential representation to obtain a three-dimensional curve frame curve corresponding to the initial image data includes: The curve potential representation is input into a decoder in the target curve processing model to obtain a three-dimensional curve frame curve corresponding to the initial image data.

5. The method according to claim 4, characterized in that The obtaining of the third training curve frame curve includes: Get the initial curve frame curve; Based on a preset starting point and a preset end point, the starting point and the end point of each curve in the initial curve frame curve are aligned to obtain a standard curve frame curve; The standard curve frame curve is discretized to obtain a third training curve frame curve.

6. The method according to claim 1, characterized in that The method further comprises: Obtaining a second training structural feature and a second training curve potential representation corresponding to the second training three-dimensional curve frame; fusing the second training structural features and the second training curve potential representation to obtain a second training curve box potential representation corresponding to the second training three-dimensional curve box; Inputting the second training curve box potential representation and the sampled noise data into an initial distribution mapping model to obtain a prediction curve box potential representation corresponding to the second training three-dimensional curve box; Based on the second training curve box potential representation and the prediction curve box potential representation, adjusting the model parameters of the initial distribution mapping model until a convergence condition is met to obtain a target distribution mapping model; The method of performing distribution mapping on the sampled noise data based on the initial image features to obtain a three-dimensional curve box potential representation corresponding to the initial image data includes: The initial image features and the sampled noise data are input into the target distribution mapping model to obtain a three-dimensional curve box potential representation corresponding to the initial image data.

7. A three-dimensional curve frame generating device, characterized in that: The device comprises: An acquisition module, used for acquiring initial image data; An encoding module, used for performing feature encoding on the initial image data to obtain initial image features corresponding to the initial image data; A distribution mapping module, used for performing distribution mapping on the sampled noise data based on the initial image features to obtain a three-dimensional curve box potential representation corresponding to the initial image data; A structure decoding module, used for performing feature decoding on the three-dimensional curve box potential representation to obtain the three-dimensional curve box structure and curve potential representation corresponding to the initial image data; A curve decoding module, used for performing feature decoding on the curve potential representation to obtain a three-dimensional curve frame curve corresponding to the initial image data; The wireframe generation module is used to generate a target three-dimensional curve frame corresponding to the initial image data based on the three-dimensional curve frame structure and the three-dimensional curve frame curve.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

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

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images