Two-dimensional garment making layout generation method and device and computer equipment

Abstract

The invention relates to a two-dimensional garment making layout generation method and device and computer equipment, and the method comprises the steps: carrying out the three-dimensional reconstruction of an input image of a target garment, and obtaining a three-dimensional model of the target garment; identifying the three-dimensional model to obtain a plurality of three-dimensional curved surface cutting pieces in the three-dimensional model; mapping each three-dimensional curved surface cutting piece into a corresponding two-dimensional tiled cutting piece, and determining cutting piece parameter information of each two-dimensional tiled cutting piece; and based on each two-dimensional tiled cutting piece and the cutting piece parameter information of each two-dimensional tiled cutting piece, generating a two-dimensional layout of the target garment. According to the method and the device, the problems of low generation efficiency of the layout and insufficient precision of the layout are solved, and the generation efficiency of the layout and the precision of the layout are improved.

Description

Technical Field

[0001] This application relates to the technical field of clothing pattern making, and particularly to a method, apparatus, and computer device for generating two-dimensional clothing pattern drawings. Background Art

[0002] A clothing pattern drawing is a drawing used to express the style structure, size specifications, and process requirements of a clothing during the clothing production and manufacturing process, and is usually used as an important technical basis for fabric cutting, sewing, and other processes. However, in the existing pattern making technology, a clothing CAD software is used to generate the pattern drawing. This method relies on manual input of pattern piece parameters and empirical design, resulting in low efficiency in generating the pattern drawing and insufficient accuracy of the pattern drawing.

[0003] Regarding the problems of low efficiency in generating the pattern drawing and insufficient accuracy of the pattern drawing in the related art, no effective solution has been proposed yet. Summary of the Invention

[0004] In this embodiment, a method, apparatus, and computer device for generating a two-dimensional clothing pattern drawing are provided to solve the problems of low efficiency in generating the pattern drawing and insufficient accuracy of the pattern drawing in the related art.

[0005] In a first aspect, in this embodiment, a method for generating a two-dimensional clothing pattern drawing is provided, including:

[0006] Performing three-dimensional reconstruction on the input image of the target clothing to obtain a three-dimensional model of the target clothing;

[0007] Identifying the three-dimensional model to obtain a plurality of three-dimensional curved surface pattern pieces in the three-dimensional model;

[0008] Mapping each of the three-dimensional curved surface pattern pieces to a corresponding two-dimensional flat pattern piece, and determining the pattern piece parameter information of each of the two-dimensional flat pattern pieces;

[0009] Generating a two-dimensional pattern drawing of the target clothing based on each of the two-dimensional flat pattern pieces and the pattern piece parameter information of each of the two-dimensional flat pattern pieces.

[0010] In some of the embodiments, the identifying the three-dimensional model to obtain a plurality of three-dimensional curved surface pattern pieces in the three-dimensional model includes:

[0011] Identifying the seam lines in the three-dimensional model through a semantic segmentation network;

[0012] Based on the recognition result, dividing the three-dimensional model into a plurality of the three-dimensional curved surface pattern pieces.

[0013] In some of the embodiments, the mapping each of the three-dimensional curved surface pattern pieces to a corresponding two-dimensional flat pattern piece includes:

[0014] Through the surface parameterization algorithm, each of the three-dimensional surface pieces is flattened into the corresponding two-dimensional tiled piece.

[0015] In some embodiments, before generating the two-dimensional pattern layout of the target garment, the method further includes:

[0016] Identifying other structural regions in the three-dimensional model and mapping the other structural regions to the corresponding two-dimensional tiled pieces; the other structural regions include pleat regions, pocket regions, and appliqué regions.

[0017] In some embodiments, the other structural region is the pleat region; the identifying other structural regions in the three-dimensional model and mapping the other structural regions to the corresponding two-dimensional tiled pieces includes:

[0018] Performing curvature analysis on the three-dimensional model to identify the pleat regions in the three-dimensional model; the curvature values of the pleat regions reach a preset curvature threshold;

[0019] Based on the curvature values of the pleat regions, mapping the pleat regions to shaded regions in the corresponding two-dimensional tiled pieces.

[0020] In some embodiments, the input image is a single-view image; performing three-dimensional reconstruction on the input image of the target garment to obtain the three-dimensional model of the target garment includes:

[0021] Extracting features from the input image of the target garment to obtain first image features of the input image;

[0022] Mapping the extracted first image features to a neural implicit field to obtain corresponding first implicit geometric representations;

[0023] Matching the first implicit geometric representations with a preset implicit geometric representation template;

[0024] Generating the three-dimensional model of the target garment based on the implicit geometric representation template that matches the first implicit geometric representations.

[0025] In some embodiments, the input image is a multi-view image; performing three-dimensional reconstruction on the input image of the target garment to obtain the three-dimensional model of the target garment includes:

[0026] Extracting features from the input image of the target garment to obtain second image features of each of the input images;

[0027] Mapping the extracted second image features to a neural implicit field to obtain corresponding second implicit geometric representations;

[0028] Optimize the second implicit geometric representation based on the photometric consistency constraint of the multi-view images.

[0029] Generate the three-dimensional model of the target clothing based on the optimized second implicit geometric representation.

[0030] In some embodiments, after generating the two-dimensional pattern drawing of the target clothing, the method further includes:

[0031] Determine the corresponding user size parameters according to the user size information.

[0032] Adjust the two-dimensional pattern drawing of the target clothing according to the user size parameters to obtain the user two-dimensional pattern drawing of the user clothing.

[0033] Generate a user three-dimensional fitting diagram based on the user size information and the user two-dimensional pattern drawing.

[0034] In some embodiments, before performing three-dimensional reconstruction on the input image of the target clothing, the method further includes:

[0035] Identify the fabric features of the target clothing through the input image, where the fabric features include at least one of the following: fabric raw material, fabric color, and fabric texture.

[0036] In some embodiments, after obtaining the user two-dimensional pattern drawing of the user clothing, the method further includes:

[0037] Train a pattern drawing optimization model based on clothing optimization information, the size optimization parameters and fabric optimization parameters corresponding to the clothing optimization information, to obtain the trained pattern drawing optimization model; the clothing optimization information includes the user size information, the fabric features, the user two-dimensional pattern drawing, and user demand information.

[0038] Adjust and optimize the user two-dimensional pattern drawing through the trained pattern drawing optimization model to obtain the optimized user two-dimensional pattern drawing.

[0039] In some embodiments, the method further includes:

[0040] Based on the three-dimensional model and fabric features of the target clothing, obtain the clothing feature information of the target clothing through a trained style-property joint model; the clothing feature information includes clothing property features and clothing style features.

[0041] In some embodiments, after obtaining the user two-dimensional pattern drawing of the user clothing, the method further includes:

[0042] Determine the area of each user's two-dimensional tiled cut piece in the user's two-dimensional layout drawing;

[0043] Based on the area of each user's two-dimensional tiled cut piece, the fabric characteristics, and the clothing characteristic information, determine the fabric consumption of the user's clothing.

[0044] In some embodiments, the method further includes:

[0045] Dynamically adjust a preset curvature threshold based on the clothing characteristic information, and identify the pleated areas based on the adjusted preset curvature threshold.

[0046] In a second aspect, a two-dimensional clothing layout drawing generation device is provided in this embodiment, including:

[0047] A reconstruction module for performing three-dimensional reconstruction on an input image of a target clothing to obtain a three-dimensional model of the target clothing;

[0048] A segmentation module for identifying the three-dimensional model to obtain a plurality of three-dimensional curved surface cut pieces in the three-dimensional model;

[0049] A flattening module for mapping each three-dimensional curved surface cut piece to a corresponding two-dimensional tiled cut piece, and determining the cut piece parameter information of each two-dimensional tiled cut piece;

[0050] A generation module for generating a two-dimensional layout drawing of the target clothing based on each two-dimensional tiled cut piece and the cut piece parameter information of each two-dimensional tiled cut piece.

[0051] In a third aspect, a computer device is provided in this embodiment, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, the two-dimensional clothing layout drawing generation method described in the first aspect above is implemented.

[0052] Compared with the related art, the two-dimensional clothing layout drawing generation method, device, and computer device provided in this embodiment perform three-dimensional reconstruction on an input image of a target clothing to obtain a three-dimensional model of the target clothing; identify the three-dimensional model to obtain a plurality of three-dimensional curved surface cut pieces in the three-dimensional model; map each three-dimensional curved surface cut piece to a corresponding two-dimensional tiled cut piece, and determine the cut piece parameter information of each two-dimensional tiled cut piece; generate a two-dimensional layout drawing of the target clothing based on each two-dimensional tiled cut piece and the cut piece parameter information of each two-dimensional tiled cut piece, solving the problems of low efficiency in generating the layout drawing and insufficient accuracy of the layout drawing, and achieving the improvement of the efficiency and accuracy of generating the layout drawing.

[0053] Details of one or more embodiments of the present application are set forth in the following drawings and description to make other features, objects, and advantages of the present application more concise and understandable. Description of the Drawings

[0054] The drawings described herein are provided to further understand the present application and form a part of the present application. The illustrative embodiments and descriptions thereof of the present application are used to explain the present application and do not constitute an improper limitation to the present application. In the drawings:

[0055] Figure 1 is a hardware structural block diagram of a terminal device for a two-dimensional clothing pattern generation method provided by an embodiment of the present application;

[0056] Figure 2 is a flowchart of a two-dimensional clothing pattern generation method provided by an embodiment of the present application;

[0057] Figure 3 is a flowchart of a three-dimensional curved surface cutting piece recognition method provided by an embodiment of the present application;

[0058] Figure 4 is a flowchart of a method for recognizing other structural areas provided by an embodiment of the present application;

[0059] Figure 5 is a flowchart of a three-dimensional reconstruction method based on a single-view image provided by an embodiment of the present application;

[0060] Figure 6 is a flowchart of a three-dimensional reconstruction method based on multi-view images provided by an embodiment of the present application;

[0061] Figure 7 is a flowchart of a two-dimensional clothing pattern optimization method provided by an embodiment of the present application;

[0062] Figure 8 is a flowchart of a two-dimensional clothing pattern generation method provided by a preferred embodiment of the present application;

[0063] Figure 9 is a structural block diagram of a two-dimensional clothing pattern generation device provided by an embodiment of the present application.

[0064] In the figure: 102, processor; 104, memory; 106, transmission device; 108, input / output device; 10, reconstruction module; 20, segmentation module; 30, flattening module; 40, generation module. Detailed Embodiments

[0065] To more clearly understand the purpose, technical solution, and advantages of the present application, the present application will be described and illustrated below with reference to the drawings and embodiments.

[0066] Unless otherwise defined, the technical terms or scientific terms involved in this application shall have the general meanings understood by those with ordinary skills in the technical field to which this application belongs. In this application, words such as "a", "one", "a kind of", "the", "these" and the like do not indicate a limitation in quantity, and they can be singular or plural. The terms "including", "comprising", "having" and any variants thereof involved in this application are intended to cover non-exclusive inclusion; for example, a process, method, system, product or device comprising a series of steps or modules (units) is not limited to the listed steps or modules (units), but may include unlisted steps or modules (units), or may include other steps or modules (units) inherent in these processes, methods, products or devices. The terms "connected", "coupled" and the like involved in this application are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The "plurality" involved in this application means two or more. "And / or" describes the association relationship of associated objects, indicating that three relationships may exist. For example, "A and / or B" may mean: A exists alone, A and B exist simultaneously, and B exists alone. Usually, the character " / " indicates that the objects associated before and after are in an "or" relationship. The terms "first", "second", "third" and the like involved in this application are only used to distinguish similar objects and do not represent a specific order for the objects.

[0067] The method embodiments provided in this embodiment can be executed on a terminal, a computer or a similar computing device. For example, when running on a terminal, Figure 1 is a hardware structure block diagram of the terminal of the two-dimensional clothing pattern generation method in this embodiment. As Figure 1 shown, the terminal may include one or more ( Figure 1 only one is shown in the figure) processors 102 and a memory 104 for storing data. Among them, the processor 102 may include, but is not limited to, a processing device such as a microprocessor MCU or a programmable logic device FPGA. The above terminal may further include a transmission device 106 for communication functions and an input / output device 108. Those of ordinary skill in the art can understand that Figure 1 the structure shown is only schematic and does not limit the structure of the above terminal. For example, the terminal may further include more or fewer components than Figure 1 shown in the figure, or have a different configuration from Figure 1 shown.

[0068] The memory 104 can be used to store computer programs, such as software programs and modules of application software, such as the computer program corresponding to the two-dimensional clothing pattern generation method in this embodiment. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, that is, the above method is implemented. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memories. In some instances, the memory 104 may further include a memory remotely disposed relative to the processor 102, and these remote memories may be connected to the terminal through a network. Examples of the above network include but are not limited to the Internet, enterprise intranet, local area network, mobile communication network, and combinations thereof.

[0069] The transmission device 106 is used to receive or send data via a network. The above network includes a wireless network provided by a communication provider of the terminal. In one instance, the transmission device 106 includes a network adapter (abbreviated as NIC), which can be connected to other network devices through a base station and thus can communicate with the Internet. In one instance, the transmission device 106 may be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0070] In this embodiment, a two-dimensional clothing pattern generation method is provided. Figure 2 is a flowchart of the two-dimensional clothing pattern generation method in this embodiment, as Figure 2 shown, this process includes the following steps:

[0071] Step S220, perform three-dimensional reconstruction on the input image of the target clothing to obtain a three-dimensional model of the target clothing;

[0072] Specifically, obtain the input image of the target clothing. The target clothing can be a short-sleeved shirt, a coat, a dress, trousers, etc. The input image is a single-view image or a multi-view image of the target clothing, and the multi-view image includes at least two images of different views. Preprocess the input image. The preprocessing operation is adapted to the type of the input image, including but not limited to image normalization processing (such as unifying the resolution, light correction) and image alignment (such as image alignment based on structure from motion or key point matching), and use algorithms such as multi-scale convolutional neural networks (such as feature pyramid network, high-resolution network), graph neural networks, etc. to extract features of the preprocessed input image to obtain the image features of the input image, and perform three-dimensional reconstruction based on the extracted image features to obtain a three-dimensional model of the target clothing. Among them, the three-dimensional reconstruction method includes a geometric-based three-dimensional reconstruction algorithm, a deep learning-based three-dimensional reconstruction algorithm, etc., which are not specifically limited here.

[0073] Exemplarily, when the input image is a single-view image of the target clothing, feature extraction is performed on the input image to obtain the first image feature of the input image. The extracted first image feature is mapped to the neural implicit field to obtain the corresponding first implicit geometric representation. Based on the first implicit geometric representation, an isosurface is extracted from the implicit field to generate a textured 3D mesh, so as to obtain the 3D model of the target clothing. In other embodiments, the first implicit geometric representation can be matched with the preset implicit geometric representation template in the clothing morphology prior library, and based on the implicit geometric representation template that matches the first implicit geometric representation, the 3D model of the target clothing is generated, thereby combining prior knowledge to optimize the 3D reconstruction process of the single-view image and improving the reconstruction efficiency and accuracy.

[0074] Exemplarily, when the input image is a multi-view image of the target clothing, feature extraction is performed on the input images of the target clothing to obtain the second image features of the input images. The extracted second image features are mapped to the neural implicit field to obtain the corresponding second implicit geometric representations. Based on the second implicit geometric representations, an isosurface is extracted from the implicit field to generate a textured 3D mesh, so as to obtain the 3D model of the target clothing. In other embodiments, the second implicit geometric representation can be optimized based on the photometric consistency constraint of the multi-view images, and based on the optimized second implicit geometric representation, the 3D model of the target clothing is generated, thereby improving the robustness of the 3D reconstruction, adapting to complex scenes, and at the same time helping to improve the reconstruction accuracy.

[0075] Furthermore, through physical engines such as Blender and Bullet, gravity and fabric flexibility constraints are applied to the 3D model mesh to dynamically adjust the shape of the clothing model through physical simulation post-processing, so as to repair unreasonable geometric distortions generated during the modeling process, such as unnatural sharp folds and excessive stretching deformations. It should be noted that applying gravity can make the 3D clothing model produce natural drooping, swinging and other motion effects that conform to the influence of real gravity in the virtual space, while applying fabric flexibility constraints is used to simulate deformation behaviors such as fabric stretching, bending and wrinkling.

[0076] In this embodiment, a single-view image or a multi-view image can be used to implement 3D reconstruction, supporting flexible input types, reducing the data acquisition threshold, and helping to improve the operation convenience.

[0077] Step S240, identify the 3D model to obtain multiple 3D curved surface pieces in the 3D model;

[0078] Specifically, the 3D model of the target clothing is identified through relevant algorithms to segment and obtain multiple 3D curved surface pieces in the 3D model. Clothing pieces are the basic units that make up a complete clothing, such as the front piece, back piece, sleeves, skirt piece, front trouser piece and back trouser piece, etc.

[0079] Among them, a segmentation algorithm based on geometric features (such as segmenting at the curvature mutation points on the model surface as the connecting boundary of the cut pieces) or a segmentation algorithm based on semantic priors can be adopted. For example, the seam lines in the three-dimensional model are identified through a semantic segmentation network, and based on the positions of the identified seam lines, the three-dimensional model is segmented into multiple three-dimensional surface cut pieces.

[0080] Step S260: Map each three-dimensional surface cut piece to a corresponding two-dimensional flat cut piece, and determine the cut piece parameter information of each two-dimensional flat cut piece.

[0081] Specifically, map each three-dimensional surface cut piece to a corresponding two-dimensional flat cut piece, and at the same time calculate and output the vertex coordinates and boundary parameters of the two-dimensional flat cut piece. The boundary parameters are used to describe the geometric features and constraint conditions of the cut piece edge, such as the boundary length, boundary curvature, and boundary curve equation, etc., to determine the basic geometric shape and key dimension information of the two-dimensional flat cut piece.

[0082] In this embodiment, multiple methods can be adopted to realize the mapping of the three-dimensional surface cut piece. For example, a surface parameterization algorithm (such as a minimum stretching parameterization algorithm, a parameterization algorithm based on conformal mapping) is used to unfold each three-dimensional surface cut piece into a corresponding two-dimensional flat cut piece; or, through a physical engine to simulate the relaxation process of clothing fabric, each three-dimensional surface cut piece is unfolded into a corresponding two-dimensional flat cut piece, while minimizing deformations such as stretching and wrinkling after unfolding; or, through geometric transformation methods such as orthogonal projection and perspective projection, each three-dimensional surface cut piece is projected into a corresponding two-dimensional flat cut piece, and the projection distortion is corrected.

[0083] Step S280: Generate a two-dimensional pattern of the target clothing based on each two-dimensional flat cut piece and the cut piece parameter information of each two-dimensional flat cut piece.

[0084] Specifically, check and correct the cut piece parameter information of each two-dimensional flat cut piece to avoid missing or contradictory parameters. Based on the checked and corrected cut piece parameter information, arrange and combine each two-dimensional flat cut piece according to preset requirements to reasonably handle the splicing relationship and alignment marks between the cut pieces. On this basis, generate an appropriate sewing allowance (which can be dynamically adjusted according to fabric characteristics and process requirements) for the edge parts that need to be sewn for each two-dimensional flat cut piece, and label each two-dimensional flat cut piece, including cut piece type labeling, positioning labeling (such as notches, drill holes) on the cut piece edge or inside, and range identification of the sewing allowance, and finally generate a complete and production-standard-compliant two-dimensional pattern of the clothing, providing an accurate technical basis for subsequent cutting and sewing processes.

[0085] A clothing pattern drawing is a drawing used to express the style structure, size specifications, and process requirements of clothing during the clothing production and manufacturing process. It is usually used as an important technical basis for fabric cutting, sewing, and other processes. However, in the existing pattern-making technology, clothing CAD software is used to generate pattern drawings. This method relies on manual input of pattern piece parameters and empirical design, resulting in low efficiency in generating pattern drawings and insufficient accuracy of pattern drawings.

[0086] Compared with the existing technology, the present application provides a method, device, and computer device for generating two-dimensional clothing pattern drawings. By performing three-dimensional reconstruction on the input image of the target clothing, a three-dimensional model of the target clothing is obtained; the three-dimensional model is recognized to obtain multiple three-dimensional curved surface pattern pieces in the three-dimensional model; each three-dimensional curved surface pattern piece is mapped to a corresponding two-dimensional flat pattern piece, and the pattern piece parameter information of each two-dimensional flat pattern piece is determined; based on each two-dimensional flat pattern piece and the pattern piece parameter information of each two-dimensional flat pattern piece, a two-dimensional pattern drawing of the target clothing is generated. Based on this, the three-dimensional clothing model can reveal the three-dimensional details that cannot be presented by the two-dimensional clothing image. By mapping each three-dimensional pattern piece in the three-dimensional clothing model to a corresponding two-dimensional flat pattern piece, accurate mapping from three-dimensional to two-dimensional is achieved, significantly improving the reduction degree of the pattern piece structure, being more in line with the physical rules of clothing, and automatically generating the two-dimensional pattern drawing corresponding to each two-dimensional flat pattern piece, optimizing the ambiguity of pattern making, and completing accurate pattern making. Thus, the problems of low efficiency in generating pattern drawings and insufficient accuracy of pattern drawings are solved, and the efficiency of generating pattern drawings and the accuracy of pattern drawings are improved, providing an accurate structural reference for subsequent production processes.

[0087] In some of these embodiments, as Figure 3 shown, the recognition of the three-dimensional model in step S240 to obtain multiple three-dimensional curved surface pattern pieces in the three-dimensional model includes the following steps:

[0088] Step S241, identifying the seam lines in the three-dimensional model through a semantic segmentation network;

[0089] Step S242, based on the recognition result, dividing the three-dimensional model into multiple three-dimensional curved surface pattern pieces.

[0090] Specifically, the seam lines on the surface mesh of the three-dimensional model are identified through a semantic segmentation network to obtain the positions of each seam line in the three-dimensional model, and based on the positions of each seam line, the three-dimensional model is divided into multiple three-dimensional curved surface pattern pieces. Clothing pattern pieces are the basic units that make up a complete piece of clothing, such as the front body piece, back body piece, sleeves, skirt pieces, front trouser pieces, and back trouser pieces, etc.

[0091] Among them, the semantic segmentation network includes but is not limited to Mask2Former, U-Net, and Mask R-CNN.

[0092] In this embodiment, the seam lines in the 3D model are identified through a semantic segmentation network. Based on the recognition results, the 3D model is segmented into multiple 3D surface pieces, achieving precise segmentation of independent garment pieces.

[0093] In some of these embodiments, mapping each 3D surface piece to a corresponding 2D flat piece in step S260 includes the following steps:

[0094] Each 3D surface piece is flattened into a corresponding 2D flat piece through a surface parameterization algorithm.

[0095] In this embodiment, surface parameterization algorithms such as the minimum stretching parameterization algorithm (such as the low-distortion parameterization method ABF++ based on angle optimization) and the parameterization algorithm based on conformal mapping are used to flatten each 3D surface piece into a corresponding 2D flat piece.

[0096] Exemplarily, the minimum stretching parameterization algorithm is applied to each 3D surface piece to flatten each 3D surface piece into a 2D polygon to obtain the corresponding 2D flat piece, ensuring that the shape after unfolding has an error less than a preset threshold, such as 2%, reducing the shape distortion during the unfolding of the piece, and making the 2D pattern highly match the garment prototype.

[0097] In this embodiment, each 3D surface piece is flattened into a corresponding 2D flat piece through a surface parameterization algorithm, achieving a fidelity mapping from the 3D surface to the 2D plane, which can effectively improve the standardization degree of garment process planning.

[0098] In some of these embodiments, before generating the 2D pattern of the target garment, the above 2D garment pattern generation method further includes the following steps:

[0099] Identify other structural regions in the 3D model and map the other structural regions to the corresponding 2D flat pieces; the other structural regions include pleat regions, pocket regions, and patch regions.

[0100] Specifically, the 3D model of the target garment is identified to obtain other structural regions in the 3D model. The other structural regions include pleat regions, pocket regions, patch regions, etc., and the other structural regions are mapped to the corresponding 2D flat pieces.

[0101] Among them, when identifying the pleat region, by performing curvature analysis on the 3D model, the region where the curvature value reaches the preset curvature threshold is identified as the pleat region in the 3D model, and based on the curvature value of the pleat region, the pleat region is mapped to the corresponding 2D flat piece.

[0102] Among them, when identifying the pocket area and the patch area, the region growing algorithm can be used, with the normalized Gaussian curvature and the normal vector consistency as the growth basis, and diffusion growth is carried out from the selected seed points to form a preliminary structural region range. Then, combined with area filtering and convexity verification, the interference regions are excluded to ensure that the identified regions are the pocket area and the patch area. Furthermore, through the conformal mapping technology, an accurate correspondence between the three-dimensional grid vertices and the two-dimensional UV coordinates is established to realize the accurate mapping of the structural region to the two-dimensional flat pattern. In other embodiments, by analyzing the Gaussian curvature of the surface mesh of the three-dimensional model, the isolated convex regions are located, and the positioning and identification of the pocket area can also be realized.

[0103] It should be noted that in the two-dimensional flat pattern, various methods can be used to mark each other structural region to clearly distinguish different structural regions. For example, different other structural regions are identified with different colors, or a unique contour line style is generated for each other structural region, or different other structural regions use different marking symbols, etc., which are not specifically limited here.

[0104] Through this embodiment, different types of other structural regions in the three-dimensional model are identified, and each other structural region is mapped to the corresponding two-dimensional flat pattern, so as to accurately identify and map different other structural regions, which helps to improve the accuracy of the subsequent generated two-dimensional pattern.

[0105] In some of these embodiments, the other structural region is a pleat region; as Figure 4 shown, identifying the other structural regions in the three-dimensional model and mapping the other structural regions to the corresponding two-dimensional flat pattern includes the following steps:

[0106] Step S251, perform curvature analysis on the three-dimensional model to identify the pleat regions in the three-dimensional model; the curvature values of the pleat regions reach the preset curvature threshold;

[0107] Step S252, based on the curvature values of the pleat regions, map the pleat regions to the shaded regions in the corresponding two-dimensional flat pattern.

[0108] Specifically, during the generation process of the two-dimensional pattern, curvature analysis is performed on the three-dimensional model of the target garment to calculate the curvature information of the surface mesh of the three-dimensional model. The curvature information includes Gaussian curvature, mean curvature, etc. The curvature information of the surface mesh of the three-dimensional model is compared with the preset curvature threshold, and the regions with curvature values reaching the preset curvature threshold are marked as pleat regions. Among them, the preset curvature threshold is set according to actual requirements such as the type of garment and the characteristics of the garment fabric.

[0109] It can be understood that points with a relatively large absolute value of Gaussian curvature usually appear at the tips or intersection areas of the folds, while the absolute value of the mean curvature is relatively large at the ridge lines or valley lines of the folds. By combining the Gaussian curvature and the mean curvature for judgment, the fold area can be accurately determined.

[0110] Furthermore, based on the curvature values of the fold area, the fold area is mapped to the corresponding shaded area in the two-dimensional tiled slice. The curvature value corresponds to the pixel grayscale of the shaded area. For example, a high brightness is used for the fold convex area, and a low brightness is used for the fold concave area, etc., to achieve the visualization of the fold area. At the same time, it supports the user to interactively adjust the shadow sensitivity, that is, by adjusting the parameters of the mapping function to change the mapping rule between the curvature value and the grayscale value to meet the analysis requirements of different precisions.

[0111] Through this embodiment, curvature analysis is performed on the three-dimensional model to identify the fold area in the three-dimensional model; the curvature value of the fold area reaches the preset curvature threshold. Based on the curvature value of the fold area, the fold area is mapped to the corresponding shaded area in the two-dimensional tiled slice, thereby realizing fold recognition and calibration, improving the structural accuracy of the two-dimensional tiled slice, and providing an accurate structural reference for the subsequent production process.

[0112] In some of these embodiments, the input image is a single-view image; as Figure 5 shown, the three-dimensional reconstruction of the input image of the target clothing in step S220 to obtain the three-dimensional model of the target clothing includes the following steps:

[0113] Step S221, extract features from the input image of the target clothing to obtain the first image feature of the input image;

[0114] Step S222, map the extracted first image feature to the neural implicit field to obtain the corresponding first implicit geometric representation;

[0115] Step S223, match the first implicit geometric representation with the preset implicit geometric representation template;

[0116] Step S224, generate the three-dimensional model of the target clothing based on the implicit geometric representation template that matches the first implicit geometric representation.

[0117] Specifically, when the input image is a single-view image, algorithms such as multi-scale convolutional neural networks and graph neural networks are used to extract features from the input image to obtain the first image feature of the input image. The first image feature includes local detail features and global semantic features.

[0118] Further, map the extracted first image features to a neural implicit field to predict the occupancy probability or the Signed Distance Function (SDF) of each point in space, so as to obtain the corresponding first implicit geometric representation. Match the first implicit geometric representation with the preset implicit geometric representation template in the clothing shape prior library, so that the implicit field prediction can be optimized by retrieving similar templates. Among them, the clothing shape prior library includes templates of different clothing types, such as the SDF templates of short sleeves and shirts.

[0119] After that, based on the implicit geometric representation template that matches the first implicit geometric representation, generate a textured 3D mesh by extracting an isosurface from the implicit field to obtain the 3D model of the target clothing.

[0120] Through this embodiment, feature extraction is performed on the input image of the target clothing to obtain the first image features of the input image. The extracted first image features are mapped to a neural implicit field to obtain the corresponding first implicit geometric representation, and the first implicit geometric representation is matched with the preset implicit geometric representation template. Based on the implicit geometric representation template that matches the first implicit geometric representation, a 3D model of the target clothing is generated, so as to optimize the 3D reconstruction process of a single-view image in combination with prior knowledge and improve the reconstruction efficiency and accuracy.

[0121] In some of these embodiments, the input image is a multi-view image; as Figure 6 shown, the three-dimensional reconstruction of the input image of the target clothing in step S220 to obtain the three-dimensional model of the target clothing includes the following steps:

[0122] Step S225, perform feature extraction on the input image of the target clothing to obtain the second image features of each input image;

[0123] Step S226, map the extracted second image features to a neural implicit field to obtain the corresponding second implicit geometric representation;

[0124] Step S227, optimize the second implicit geometric representation based on the photometric consistency constraint of the multi-view image;

[0125] Step S228, generate the three-dimensional model of the target clothing based on the optimized second implicit geometric representation.

[0126] Specifically, when the input image is a multi-view image, algorithms such as a multi-scale convolutional neural network and a graph neural network are used to perform feature extraction on the input image to obtain the second image features of the input image. For example, the multi-view image is input into an encoder with shared weights, multiple groups of image features are extracted, and the multiple groups of image features are fused through a cross-view attention mechanism (such as a Transformer architecture), and the fused second image features are output.

[0127] Further, map the extracted second image features to a neural implicit field to predict the occupancy probability or signed distance function of each point in space, so as to obtain the corresponding second implicit geometric representation, and optimize the second implicit geometric representation based on the photometric consistency constraint of multi-view images. Among them, use differentiable rendering (such as PyTorch3D, TensorFlowGraphics) to project the implicit field onto each view to obtain a rendered image, compare the rendered image with the real multi-view image pixel by pixel, calculate the loss value through a photometric loss function, and adjust the expression of the implicit field based on the calculated photometric consistency loss to reduce the difference between the rendered image and the real image, ensure the lighting rationality, and help to achieve fine optimization of the details of the 3D model in a multi-view scenario.

[0128] After that, based on the optimized second implicit geometric representation, generate a textured 3D mesh by extracting an isosurface from the implicit field to obtain a 3D model of the target clothing.

[0129] Through this embodiment, feature extraction is performed on the input images of the target clothing to obtain the second image features of each input image, map the extracted second image features to a neural implicit field to obtain the corresponding second implicit geometric representation, optimize the second implicit geometric representation based on the photometric consistency constraint of multi-view images, and then generate a 3D model of the target clothing based on the optimized second implicit geometric representation, thereby improving the robustness of 3D reconstruction, adapting to complex scenarios, and at the same time helping to improve the reconstruction accuracy.

[0130] In some of these embodiments, as Figure 7 shown, after generating the 2D pattern drawing of the target clothing, the following steps are further included:

[0131] Step S291: Determine the corresponding user size parameters according to the user size information;

[0132] Step S292: Adjust the 2D pattern drawing of the target clothing according to the user size parameters to obtain the user 2D pattern drawing of the user clothing;

[0133] Step S293: Generate a user 3D fitting diagram based on the user size information and the user 2D pattern drawing.

[0134] Specifically, user size information is obtained. The user size information can be in the form of an image (such as a full-body image or a half-body image of the user) and / or in the form of text (such as key size parameters like chest circumference, waist circumference, hip circumference, shoulder width, etc., or user body type descriptions like being thin or having a pear-shaped body). When the user size information is user size parameters, the user size parameters are directly obtained. When the user size information is an image or a body type description, through a size parameter model, the user size information is converted into corresponding user size parameters. Among them, the size parameter model is trained through user body type images, body type descriptions, and user size parameters.

[0135] Based on the user size parameters, the two-dimensional pattern is adjusted through a parametric mapping relationship to obtain the user's two-dimensional pattern of the clothing, that is, the user's two-dimensional pattern adapted to the user's body shape. Among them, in advance, through parametric mapping rules, the user size is associated with the key control points of the pieces in the two-dimensional pattern (such as the inflection points of the neckline arc and the endpoints of the darts). Through a non-uniform scaling algorithm, differential displacement calculations are performed on each control point to dynamically adjust the contour of the piece. At the same time, a constraint solving algorithm is used to maintain the topological structure of the piece, and finally, the user's two-dimensional pattern adapted to the user's body shape is generated.

[0136] Furthermore, according to the user size information, a user image model is generated. For example, a human body model is reconstructed through 3D scanning or multi-view photography, or a parametric mannequin template is driven by input size data, such as a deformable human body model (Skinned Multi-Person Linear Model, SMPL model), to generate a 3D human body mesh that fits the user's size, or a standard mannequin image is directly called.

[0137] In this embodiment, virtual fitting can be performed based on the user image modeling result. Specifically, according to the user's two-dimensional pattern, a corresponding 3D image of the user's clothing is produced, and through a virtual fitting tool, the clothing model is matched with the user image model to achieve the display of the virtual fitting effect, or the pieces in the user's two-dimensional pattern are adsorbed to the corresponding positions on the 3D human body according to the actual sewing logic (such as aligning the sleeves with the shoulders). An accurate spatial correspondence relationship between the pieces and the human body surface is established through key point projection or UV mapping. Subsequently, based on a physics engine, the sewing constraints between the pieces are simulated, and fabric mechanical properties (such as elasticity and drapability) are applied to simulate the natural form of the clothing when worn. Finally, a 3D fitting diagram is output. In this way, a combination scheme of piece spatial positioning and virtual sewing is adopted to restore the real sewing logic and improve the credibility of the fitting.

[0138] Through this embodiment, according to the user's size information, the corresponding user size parameters are determined, and the two-dimensional pattern of the target clothing is adjusted according to the user size parameters to obtain the user's two-dimensional pattern of the clothing, realizing the precise adaptation of the clothing pattern to the human body shape. Based on the user size information and the user's two-dimensional pattern, a user's three-dimensional fitting image is generated to realize the visual verification of clothing design. Furthermore, the user's two-dimensional pattern is further adjusted based on the fitting effect to improve the accuracy of the clothing pattern.

[0139] In some of these embodiments, before performing three-dimensional reconstruction on the input image of the target clothing, the above two-dimensional clothing pattern generation method further includes the following steps:

[0140] Identify the fabric characteristics of the target clothing through the input image, where the fabric characteristics include at least one of the following: fabric raw material, fabric color, fabric texture.

[0141] Specifically, the input image of the target clothing is identified to obtain the fabric characteristics of the target clothing, which include fabric raw material, fabric color, fabric texture, etc. Among them, the fabric raw material is the basic material that constitutes the fabric, including natural fibers (such as cotton, linen), chemical fibers (such as rayon, polyester, nylon), etc. The fabric texture refers to the arrangement and combination of fibers or yarns that make up the fabric, including yarn structure (such as single yarn, ply yarn), fabric structure (such as plain weave, twill weave, satin weave), etc.

[0142] It can be understood that the implementation method of the fabric characteristic recognition process can be selected according to actual application requirements, including but not limited to image processing algorithms, machine learning models, or a combination thereof, and will not be specifically limited here.

[0143] Exemplarily, for fabric raw material recognition, texture features such as contrast, energy, and entropy can be extracted through the gray-level co-occurrence matrix to analyze the microscopic details of fiber texture in the input image to determine the fabric raw material used for the target clothing. For example, cotton fabrics usually have rough textures, and silk has smooth textures; or, a Gabor filter can be used to capture the fiber arrangement pattern (such as chemical fibers having a regular grid texture) through multi-directional and multi-scale filtering responses to determine the fabric raw material used for the target clothing; or, high-dimensional texture features can be extracted through a pre-trained model (such as a residual network, VGG), combined with a fine-tuned classification layer to achieve fabric type classification. At the same time, the channel attention module can be used to focus on the key areas of the fabric (such as fiber joints) to improve the fine-grained classification accuracy and accurately determine the fabric raw material used for the target clothing.

[0144] Exemplarily, for fabric color recognition, the input image is converted from the RGB color space to the HSV color space or the Lab color space, and the illumination interference is reduced by separating the luminance and chrominance information, and then the fabric color is accurately recognized based on the chrominance information; or, color histograms are statistically calculated for the chrominance channels (such as the H channel of the HSV color space, the a channel and the b channel of the Lab color space), and the accurate determination of the fabric color is achieved by matching with a preset color template (such as Pantone color card, Natural Color System color card); or, an unsupervised clustering algorithm (such as K-means clustering algorithm, density-based clustering algorithm DBSCAN) is used to extract the main color, and the color distribution can also be statistically calculated after separating the clothing area of the input image through a semantic segmentation model (such as U-Net, fully convolutional network) to accurately determine the fabric color.

[0145] Exemplarily, for fabric texture recognition, the local binary pattern is used to quantify the gray-scale difference between pixels, and the diagonal features of twill weave and other details are accurately recognized by capturing the interweaving rules of warp and weft yarns; or, the input image is converted to the frequency domain through Fourier transform, and structures such as plain weave (high-frequency energy concentrated) and satin weave (low-frequency energy dominant) are distinguished based on the energy distribution characteristics; or, object detection models such as Faster R-CNN and YOLO are used to locate fabric unit structures such as yarn intersections to achieve rapid classification and recognition of the fabric texture; or, a graph convolutional network is used to model the topological connection relationship of yarns to effectively recognize complex fabric textures such as jacquard patterns.

[0146] Through this embodiment, the fabric features of the target clothing are recognized from the input image, and the fabric features at least include one of fabric raw materials, fabric colors, and fabric textures, so as to accurately recognize the fabric features of the clothing.

[0147] In some of these embodiments, after obtaining the user's two-dimensional clothing layout of the user's clothing, the above two-dimensional clothing layout generation method further includes the following steps:

[0148] Based on the clothing optimization information, the size optimization parameters and fabric optimization parameters corresponding to the clothing optimization information, the layout optimization model is trained to obtain a trained layout optimization model; the clothing optimization information includes user size information, fabric features, the user's two-dimensional clothing layout, and user demand information;

[0149] Through the trained layout optimization model, the user's two-dimensional clothing layout is adjusted and optimized to obtain an optimized user's two-dimensional clothing layout.

[0150] Specifically, a training dataset is constructed, which includes clothing optimization information and its corresponding 3D image optimization information and fabric feature optimization information. Among them, the clothing optimization information includes user size information, fabric feature information, user's 2D pattern drawing, and user demand information. The user demand information includes user size demand information and / or user fabric demand information. The user demand information can be text and / or images, and the user demand information can be obtained by the user referring to the 3D virtual fitting image. For example, if the neckline in the 3D virtual fitting image is slightly large, the user demand information may include "reduce the neckline size". The size optimization parameter and the fabric optimization parameter are adjustment parameters for adjusting the user's 2D pattern drawing. For example, if the user demand information is "the waist curve is more fitting" and "red", the corresponding size optimization parameter is the 2D pattern drawing adjustment parameter that makes the waistline part in the 2D pattern drawing closer to the user's waist circumference, and the fabric optimization parameter is the fabric feature adjustment parameter that meets the user's color demand.

[0151] Further, based on the clothing optimization information, the corresponding size optimization parameter and fabric optimization parameter of the clothing optimization information, the pattern drawing optimization model is trained to obtain a trained pattern drawing optimization model. Then, the actual user size information, fabric feature information, user's 2D pattern drawing, and user demand information are input into the trained pattern drawing optimization model to obtain the size optimization parameter and / or fabric optimization parameter, and the user's 2D pattern drawing is adjusted according to the size optimization parameter and / or fabric optimization parameter to obtain an optimized user's 2D pattern drawing.

[0152] Through this embodiment, based on the clothing optimization information and its corresponding size optimization parameter and fabric optimization parameter, the pattern drawing optimization model is trained to obtain a trained pattern drawing optimization model, and through the trained pattern drawing optimization model, the user's 2D pattern drawing is adjusted and optimized to obtain an optimized user's 2D pattern drawing, improving the accuracy of the clothing pattern drawing, realizing the personalized generation of the clothing pattern drawing, and supporting customized production.

[0153] In some of these embodiments, the above method for generating a 2D clothing pattern drawing further includes the following steps:

[0154] Based on the 3D model of the target clothing and fabric features, through a trained style-property joint model, the clothing feature information of the target clothing is obtained; the clothing feature information includes clothing physical property features and clothing style features.

[0155] Specifically, a training dataset is constructed, which includes the 3D image of the clothing, fabric features, and their corresponding clothing feature information. The clothing feature information includes clothing physical property features and clothing style features. The clothing style features can be manually labeled, and the clothing physical property features can be obtained through measurement or simulation, and the pre-constructed style-property joint model is trained according to the training dataset.

[0156] Further, input the 3D model of the target garment and the fabric characteristics into the trained style-property joint model, perform multi-layer feature extraction and semantic parsing on the input data, and obtain the garment feature information of the target garment. The garment feature information includes garment property features and garment style features.

[0157] Among them, the garment property features include objectively measurable fabric physical and chemical properties or other objective properties, such as shrinkage rate, elasticity, etc. The garment style features include the type features of the target garment (such as casual wear, professional wear), aesthetic features (such as elegance, lively, fashionable), scene features (such as outdoor occasions, home occasions, formal occasions), appearance features (such as drapability, glossiness), and tactile features (such as soft, rough), etc. The glossiness includes forms such as mercerization, high gloss, etc.

[0158] Through this embodiment, based on the 3D model of the target garment and the fabric characteristics, the garment feature information of the target garment is obtained through the trained style-property joint model; the garment feature information includes garment property features and garment style features, thereby combining physical properties with semantic style labels, constructing a multi-modal feature space, realizing cross-modal reasoning, and thus being able to give functional and aesthetic features simultaneously, supporting customized production.

[0159] In some of these embodiments, after obtaining the user's 2D pattern of the user's garment, the following steps are further included:

[0160] Determine the area of each user's 2D flat laying cut piece in the user's 2D pattern;

[0161] Based on the area of each user's 2D flat laying cut piece, the fabric characteristics, and the garment feature information, determine the fabric consumption of the user's garment.

[0162] Specifically, according to the cut piece parameter information of each user's 2D flat laying cut piece in the user's 2D pattern of the user's garment, and the cut piece structure (such as pleat area, pocket area, patch area) in the user's 2D flat laying cut piece, calculate the area of each user's 2D flat laying cut piece, and combine the fabric characteristics and the garment feature information of the user's garment to estimate the fabric consumption of the user's garment. For example, when the user's garment uses pure cotton fabric, considering the high shrinkage rate of pure cotton fabric, a shrinkage allowance needs to be reserved when estimating the fabric consumption to achieve accurate estimation and ensure the quality of the finished product.

[0163] Further, intelligent nesting can be performed based on factors such as the width of the fabric, the texture direction, and the pattern repeat period to optimize the cut piece layout, which helps to improve the accuracy of estimating the fabric consumption and reduce production losses at the same time.

[0164] Through this embodiment, the area of each user's 2D tiled cutting piece in the user's 2D layout drawing is determined, and based on the area of each user's 2D tiled cutting piece, fabric characteristics, and clothing feature information, the fabric consumption of the user's clothing is determined, realizing fabric calculation and nesting optimization to reduce raw material costs and losses, while improving production efficiency and the finished clothing effect.

[0165] In some of these embodiments, the above 2D clothing layout drawing generation method further includes the following steps:

[0166] Dynamically adjust a preset curvature threshold based on clothing feature information, and identify a wrinkled area based on the adjusted preset curvature threshold.

[0167] Specifically, the preset curvature threshold is used to analyze and determine the wrinkled area of the target clothing. The curvature information of the surface mesh of the 3D model is compared with the preset curvature threshold, and the area where the curvature value reaches the preset curvature threshold is detected as the wrinkled area.

[0168] In this embodiment, the preset curvature threshold is dynamically adjusted according to the clothing feature information of the target clothing. For example, when the fabric used for the target clothing is high-elasticity spandex, the preset curvature threshold is dynamically lowered to adapt to the fabric of the target clothing, and the wrinkled area is identified based on the adjusted preset curvature threshold to avoid missed detection or over-detection of wrinkles.

[0169] Through this embodiment, the preset curvature threshold is dynamically adjusted based on clothing feature information, and the wrinkled area is identified based on the adjusted preset curvature threshold, realizing accurate identification of wrinkles, improving the structural accuracy of the 2D tiled cutting piece, and helping to accurately calculate the fabric consumption of the clothing subsequently.

[0170] The following describes and illustrates this embodiment through preferred embodiments.

[0171] Figure 8 is a flowchart of the 2D clothing layout drawing generation method of this preferred embodiment, as Figure 8 shown, the 2D clothing layout drawing generation method includes the following steps:

[0172] Step S801, identify the fabric characteristics of the target clothing through the input image, and the fabric characteristics include fabric raw material, fabric color, and fabric tissue structure;

[0173] Step S802, perform 3D reconstruction on the input image of the target clothing to obtain a 3D model of the target clothing;

[0174] Step S803, identify the seam lines in the 3D model through a semantic segmentation network, and based on the recognition result, divide the 3D model into multiple 3D curved surface cutting pieces;

[0175] Step S804: Flatten each 3D surface piece into a corresponding 2D tiled piece through a surface parameterization algorithm, and determine the piece parameter information of each 2D tiled piece.

[0176] Step S805: Identify other structural regions in the 3D model and map the other structural regions to the corresponding 2D tiled pieces; the other structural regions include pleat regions, pocket regions, and patch regions.

[0177] Step S806: Generate a 2D pattern layout of the target garment based on each 2D tiled piece and the piece parameter information of each 2D tiled piece.

[0178] Step S807: Determine the corresponding user size parameters according to the user size information, and adjust the 2D pattern layout of the target garment according to the user size parameters to obtain the user 2D pattern layout of the user's garment.

[0179] Step S808: Generate a user 3D fitting image based on the user size information and the user 2D pattern layout.

[0180] Step S809: Train a pattern layout optimization model based on the garment optimization information and its corresponding size optimization parameters and fabric optimization parameters to obtain a trained pattern layout optimization model; adjust and optimize the user 2D pattern layout through the trained pattern layout optimization model to obtain an optimized user 2D pattern layout.

[0181] Step S810: Based on the 3D model of the target garment and the fabric characteristics, obtain the garment feature information of the target garment through a trained style-physical property joint model; the garment feature information includes garment physical property features and garment style features.

[0182] Step S811: Determine the fabric consumption of the user's garment based on the area of each user 2D tiled piece, the fabric characteristics, and the garment feature information in the user 2D pattern layout.

[0183] Through this embodiment, the fabric characteristics of the target garment are recognized by inputting an image. The fabric characteristics include fabric raw materials, fabric colors, and fabric tissue structures. A 3D reconstruction is performed on the input image of the target garment to obtain a 3D model of the target garment. The seam lines in the 3D model are recognized through a semantic segmentation network, and based on the recognition results, the 3D model is segmented into multiple 3D surface pieces.

[0184] Further, through a surface parameterization algorithm, each three-dimensional surface piece is flattened into a corresponding two-dimensional tiled piece, the piece parameter information of each two-dimensional tiled piece is determined, and other structural regions in the three-dimensional model are identified and mapped to the corresponding two-dimensional tiled pieces. The other structural regions include pleat regions, pocket regions, and patch regions. Based on each two-dimensional tiled piece and the piece parameter information of each two-dimensional tiled piece, a two-dimensional pattern drawing of the target garment is generated. According to the user size information, the corresponding user size parameters are determined, and the two-dimensional pattern drawing of the target garment is adjusted according to the user size parameters to obtain the user two-dimensional pattern drawing of the user garment. Based on the user size information and the user two-dimensional pattern drawing, a user three-dimensional fitting diagram is generated.

[0185] Further, based on the garment optimization information and its corresponding size optimization parameters and fabric optimization parameters, the pattern drawing optimization model is trained to obtain a trained pattern drawing optimization model; through the trained pattern drawing optimization model, the user two-dimensional pattern drawing is adjusted and optimized to obtain an optimized user two-dimensional pattern drawing, improving the accuracy of the garment pattern drawing, realizing the personalized generation of the garment pattern drawing, and supporting customized production.

[0186] After that, based on the three-dimensional model and fabric characteristics of the target garment, through a trained style-property joint model, the garment characteristic information of the target garment is obtained. The garment characteristic information includes garment physical property characteristics and garment style characteristics. Based on the area, fabric characteristics, and garment characteristic information of each user two-dimensional tiled piece in the user two-dimensional pattern drawing, the fabric consumption of the user garment is determined, solving the problems of low pattern drawing generation efficiency and insufficient pattern drawing accuracy, realizing the improvement of pattern drawing generation efficiency and pattern drawing accuracy, and at the same time being able to accurately estimate the fabric consumption required for the target garment, which helps to reduce raw material costs and losses and improve production efficiency.

[0187] It should be noted that the steps shown in the above process or the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. And although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than here.

[0188] In this embodiment, a two-dimensional garment pattern drawing generation device is also provided. This device is used to implement the above embodiments and preferred implementation manners, and those that have been described will not be repeated. The following terms such as "module", "unit", "sub-unit", etc. can be a combination of software and / or hardware that can achieve a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, implementation in hardware, or a combination of software and hardware is also possible and contemplated.

[0189] Figure 9is the structural block diagram of the two-dimensional garment pattern generation device of this embodiment, as Figure 9 shown. The device includes:

[0190] A reconstruction module 10, configured to perform three-dimensional reconstruction on the input image of the target garment to obtain a three-dimensional model of the target garment;

[0191] A segmentation module 20, configured to identify the three-dimensional model to obtain multiple three-dimensional curved surface cut pieces in the three-dimensional model;

[0192] A flattening module 30, configured to map each three-dimensional curved surface cut piece to a corresponding two-dimensional tiled cut piece, and determine the cut piece parameter information of each two-dimensional tiled cut piece;

[0193] A generation module 40, configured to generate a two-dimensional pattern of the target garment based on each two-dimensional tiled cut piece and the cut piece parameter information of each two-dimensional tiled cut piece.

[0194] Through the device provided in this embodiment, by performing three-dimensional reconstruction on the input image of the target garment, a three-dimensional model of the target garment is obtained; the three-dimensional model is identified to obtain multiple three-dimensional curved surface cut pieces in the three-dimensional model; each three-dimensional curved surface cut piece is mapped to a corresponding two-dimensional tiled cut piece, and the cut piece parameter information of each two-dimensional tiled cut piece is determined; based on each two-dimensional tiled cut piece and the cut piece parameter information of each two-dimensional tiled cut piece, a two-dimensional pattern of the target garment is generated, which solves the problems of low pattern generation efficiency and insufficient pattern accuracy, and realizes the improvement of pattern generation efficiency and pattern accuracy.

[0195] In some embodiments, the segmentation module 20 is further configured to identify the seam lines in the three-dimensional model through a semantic segmentation network; based on the recognition result, the three-dimensional model is segmented into multiple three-dimensional curved surface cut pieces.

[0196] In some embodiments, the flattening module 30 is further configured to flatten each three-dimensional curved surface cut piece into a corresponding two-dimensional tiled cut piece through a surface parameterization algorithm.

[0197] In some embodiments, the flattening module 30 is further configured to identify other structural regions in the three-dimensional model, and map the other structural regions to the corresponding two-dimensional tiled cut pieces; the other structural regions include pleat regions, pocket regions, and appliqué regions.

[0198] In some embodiments, the flattening module 30 is further configured to perform curvature analysis on the three-dimensional model to identify the pleat regions in the three-dimensional model; the curvature value of the pleat regions reaches a preset curvature threshold; based on the curvature value of the pleat regions, the pleat regions are mapped to the shadow regions in the corresponding two-dimensional tiled cut pieces.

[0199] In some of these embodiments, the reconstruction module 10 is further configured to extract features from the input image of the target garment to obtain the first image features of the input image; map the extracted first image features to the neural implicit field to obtain the corresponding first implicit geometric representation; match the first implicit geometric representation with a preset implicit geometric representation template; and generate a three-dimensional model of the target garment based on the implicit geometric representation template that matches the first implicit geometric representation.

[0200] In some of these embodiments, the reconstruction module 10 is further configured to extract features from the input image of the target garment to obtain the second image features of each input image; map the extracted second image features to the neural implicit field to obtain the corresponding second implicit geometric representation; optimize the second implicit geometric representation based on the photometric consistency constraint of the multi-view images; and generate a three-dimensional model of the target garment based on the optimized second implicit geometric representation.

[0201] In some of these embodiments, Figure 9 On this basis, the device further includes an adaptation module, configured to determine the corresponding user size parameters according to the user size information; adjust the two-dimensional pattern drawing of the target garment according to the user size parameters to obtain the user two-dimensional pattern drawing of the user garment; and generate a user three-dimensional fitting image based on the user size information and the user two-dimensional pattern drawing.

[0202] In some of these embodiments, Figure 9 On this basis, the device further includes an identification module, configured to identify the fabric features of the target garment through the input image, where the fabric features include at least one of the following: fabric raw material, fabric color, and fabric texture structure.

[0203] In some of these embodiments, Figure 9 On this basis, the device further includes an optimization module, configured to train a pattern drawing optimization model based on the garment optimization information, the size optimization parameters corresponding to the garment optimization information, and the fabric optimization parameters to obtain a trained pattern drawing optimization model; the garment optimization information includes user size information, fabric features, user two-dimensional pattern drawing, and user demand information; and adjust and optimize the user two-dimensional pattern drawing through the trained pattern drawing optimization model to obtain an optimized user two-dimensional pattern drawing.

[0204] In some of these embodiments, Figure 9 On this basis, the device further includes an extraction module, configured to obtain the garment feature information of the target garment through a trained style-property joint model based on the three-dimensional model of the target garment and the fabric features; the garment feature information includes garment property features and garment style features.

[0205] In some of these embodiments, Figure 9On this basis, the device further includes an analysis module for determining the area of each user's two-dimensional tiled cut piece in the user's two-dimensional layout drawing; and determining the fabric consumption of the user's clothing based on the area of each user's two-dimensional tiled cut piece, fabric characteristics, and clothing feature information.

[0206] In some of these embodiments, on Figure 9 On this basis, the device further includes an adjustment module for dynamically adjusting a preset curvature threshold based on clothing feature information, and identifying a wrinkled area based on the adjusted preset curvature threshold.

[0207] It should be noted that each of the above modules can be a functional module or a program module, and can be implemented either by software or by hardware. For modules implemented by hardware, each of the above modules can be located in the same processor; or each of the above modules can also be located in different processors in any combined form.

[0208] In this embodiment, a computer device is also provided, including a memory and a processor. A computer program is stored in the memory, and the processor is configured to run the computer program to execute the steps in any one of the above method embodiments.

[0209] Optionally, the above computer device may further include a transmission device and input / output devices. Among them, the transmission device is connected to the above processor, and the input / output devices are connected to the above processor.

[0210] Optionally, in this embodiment, the above processor may be configured to execute the following steps through a computer program:

[0211] S1, performing three-dimensional reconstruction on the input image of the target clothing to obtain a three-dimensional model of the target clothing;

[0212] S2, identifying the three-dimensional model to obtain multiple three-dimensional curved surface cut pieces in the three-dimensional model;

[0213] S3, mapping each three-dimensional curved surface cut piece to a corresponding two-dimensional tiled cut piece, and determining the cut piece parameter information of each two-dimensional tiled cut piece;

[0214] S4, generating a two-dimensional layout drawing of the target clothing based on each two-dimensional tiled cut piece and the cut piece parameter information of each two-dimensional tiled cut piece.

[0215] It should be noted that specific examples in this embodiment can refer to the examples described in the above embodiments and optional implementation manners, and will not be elaborated in this embodiment.

[0216] In addition, in combination with the two-dimensional garment pattern generation method provided in the above embodiments, a storage medium can also be provided in this embodiment to implement it. A computer program is stored on the storage medium; when the computer program is executed by a processor, any one of the two-dimensional garment pattern generation methods in the above embodiments is implemented.

[0217] It should be understood that the specific embodiments described here are only used to explain this application, rather than to limit it. According to the embodiments provided in this application, all other embodiments obtained by those of ordinary skill in the art without creative work belong to the protection scope of this application.

[0218] Obviously, the accompanying drawings are only some examples or embodiments of this application. For those of ordinary skill in the art, this application can also be applied to other similar situations based on these drawings without creative work. In addition, it can be understood that although the work done during this development process may be complex and time-consuming, for those of ordinary skill in the art, certain design, manufacturing, or production changes based on the technical content disclosed in this application are only conventional technical means and should not be regarded as insufficient disclosure of this application.

[0219] The term "embodiment" in this application means that the specific features, structures, or characteristics described in combination with the embodiments can be included in at least one embodiment of this application. The phrase appears in various positions in the specification and does not necessarily mean the same embodiment, nor does it mean being independent or alternative to other embodiments and mutually exclusive. Those of ordinary skill in the art can clearly or implicitly understand that the embodiments described in this application can be combined with other embodiments without conflict.

[0220] The above-described embodiments only represent several implementation manners of this application, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of patent protection. It should be noted that for those of ordinary skill in the art, without departing from the concept of this application, several modifications and improvements can still be made, and these all belong to the protection scope of this application. Therefore, the protection scope of this application should be subject to the appended claims.

Claims

1. A method for generating a two-dimensional clothing pattern, characterized in that, Including: Performing three-dimensional reconstruction on an input image of a target garment to obtain a three-dimensional model of the target garment; Identifying the three-dimensional model to obtain multiple three-dimensional curved surface cut pieces in the three-dimensional model; Mapping each of the three-dimensional curved surface cut pieces to a corresponding two-dimensional flat-laid cut piece, and determining cut piece parameter information of each of the two-dimensional flat-laid cut pieces; Generating a two-dimensional pattern drawing of the target garment based on each of the two-dimensional flat-laid cut pieces and the cut piece parameter information of each of the two-dimensional flat-laid cut pieces.

2. The method for generating a two-dimensional clothing pattern according to claim 1, wherein The identifying the three-dimensional model to obtain multiple three-dimensional curved surface cut pieces in the three-dimensional model includes: Identifying seam lines in the three-dimensional model through a semantic segmentation network; Based on the identification result, dividing the three-dimensional model into multiple three-dimensional curved surface cut pieces.

3. The method for generating a two-dimensional clothing pattern according to claim 1, characterized in that, The mapping each of the three-dimensional curved surface cut pieces to a corresponding two-dimensional flat-laid cut piece includes: Flattening each of the three-dimensional curved surface cut pieces into the corresponding two-dimensional flat-laid cut piece through a surface parameterization algorithm.

4. The method for generating a two-dimensional clothing pattern according to claim 1, wherein Before generating the two-dimensional pattern drawing of the target garment, the method further includes: Identifying other structural regions in the three-dimensional model, and mapping the other structural regions to the corresponding two-dimensional flat-laid cut pieces; the other structural regions include a pleat region, a pocket region, and a patch region.

5. The method for generating a two-dimensional clothing pattern according to claim 4, wherein The other structural region is the pleat region; The identifying other structural regions in the three-dimensional model and mapping the other structural regions to the corresponding two-dimensional flat-laid cut pieces includes: Performing curvature analysis on the three-dimensional model to identify the pleat region in the three-dimensional model; the curvature value of the pleat region reaches a preset curvature threshold; Based on the curvature value of the pleat region, mapping the pleat region to a shadow region in the corresponding two-dimensional flat-laid cut piece.

6. The method for generating a two-dimensional clothing pattern according to claim 1, wherein The input image is a single-view image; Performing three-dimensional reconstruction on an input image of a target garment to obtain a three-dimensional model of the target garment includes: Performing feature extraction on the input image of the target garment to obtain a first image feature of the input image; Mapping the extracted first image feature to a neural implicit field to obtain a corresponding first implicit geometric representation; Matching the first implicit geometric representation with a preset implicit geometric representation template; Generating the three-dimensional model of the target garment based on the implicit geometric representation template that matches the first implicit geometric representation.

7. The method for generating a two-dimensional clothing pattern according to claim 1, wherein The input image is a multi-view image; performing three-dimensional reconstruction on an input image of a target garment to obtain a three-dimensional model of the target garment includes: Performing feature extraction on the input images of the target garment to obtain second image features of the input images; Mapping the extracted second image features to a neural implicit field to obtain corresponding second implicit geometric representations; Optimizing the second implicit geometric representations based on the photometric consistency constraint of the multi-view images; Generating the three-dimensional model of the target garment based on the optimized second implicit geometric representations.

8. The method for generating a two-dimensional clothing pattern according to claim 1, wherein After generating the two-dimensional pattern drawing of the target garment, it further includes: Determining corresponding user size parameters according to user size information. Adjust the two-dimensional layout drawing of the target clothing according to the user size parameter to obtain the user two-dimensional layout drawing of the user clothing; Generate a user 3D fitting image based on the user size information and the user two-dimensional layout drawing.

9. The method for generating a two-dimensional clothing pattern according to claim 8, characterized in that Before performing 3D reconstruction on the input image of the target clothing, the method further includes: Identify the fabric characteristics of the target clothing through the input image, where the fabric characteristics include at least one of the following: fabric raw material, fabric color, and fabric tissue structure.

10. The method for generating a two-dimensional clothing pattern according to claim 9, characterized in that, After obtaining the user two-dimensional layout drawing of the user clothing, the method further includes: Train a layout drawing optimization model based on clothing optimization information, the size optimization parameter and the fabric optimization parameter corresponding to the clothing optimization information to obtain the trained layout drawing optimization model; the clothing optimization information includes the user size information, the fabric characteristics, the user two-dimensional layout drawing and user demand information; Adjust and optimize the user two-dimensional layout drawing through the trained layout drawing optimization model to obtain the optimized user two-dimensional layout drawing.

11. The method for generating a two-dimensional clothing pattern according to claim 9 or 10, characterized in that The method further includes: Based on the 3D model and fabric characteristics of the target clothing, obtain the clothing feature information of the target clothing through a trained style-physical property joint model; the clothing feature information includes clothing physical property features and clothing style features.

12. The method for generating a two-dimensional clothing pattern according to claim 11, wherein After obtaining the user two-dimensional layout drawing of the user clothing, the method further includes: Determine the area of each user two-dimensional flat cut piece in the user two-dimensional layout drawing; Determine the fabric consumption of the user clothing based on the area of each user two-dimensional flat cut piece, the fabric characteristics and the clothing feature information.

13. The method for generating a two-dimensional clothing pattern according to claim 11, wherein The method further includes: Dynamically adjust a preset curvature threshold based on the clothing feature information, and identify a wrinkled area based on the adjusted preset curvature threshold.

14. A two-dimensional clothing pattern generation device, characterized in that, Includes: A reconstruction module for performing 3D reconstruction on the input image of the target clothing to obtain the 3D model of the target clothing; A segmentation module for identifying the 3D model to obtain multiple 3D curved surface cut pieces in the 3D model; A flattening module for mapping each 3D curved surface cut piece to a corresponding 2D flat cut piece and determining the cut piece parameter information of each 2D flat cut piece; A generation module for generating the two-dimensional layout drawing of the target clothing based on each 2D flat cut piece and the cut piece parameter information of each 2D flat cut piece.

15. A computer device, comprising a memory and a processor, characterized in that, A computer program is stored in the memory, and the processor is configured to run the computer program to execute the steps of the two-dimensional clothing layout drawing generation method according to any one of claims 1 to 13.

Images