Intelligent push plate method and device for garment customization, medium and product
Through neural network model training and automated processing of model size templates, the problem of time-consuming and large errors of traditional clothing pushing boards is solved, and intelligent clothing customization is realized, which improves the efficiency and accuracy of clothing customization.
Patent Information
- Application Number
- CN202411277405.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-12
- Publication Date
- 2024-12-17
AI Technical Summary
Traditional clothing push boards rely on manual work, take a long time and have large errors, so they cannot effectively customize clothing automatically and intelligently according to the customer's body shape.
Using neural network model, we can obtain the key points of the model's net size and size templates of human body, train it to generate intelligent push board methods for clothing customization, and use the trained model to generate customized clothing templates for customers to achieve automation and intelligence. Custom clothing.
It greatly improves the efficiency of clothing customization and can provide fitted clothing for most customers. Only a few models need to try on and train models, reducing the time and error of manual adjustments.
Smart Images
Figure CN119151639A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of digital clothing, and in particular to an intelligent push plate method, device, medium and product for clothing customization. Background Art
[0002] Wearing clothes should be comfortable and fit, and due to individual differences, people's height, weight, and body shape are all different, which requires that clothes should meet the needs of different body types. Clothing push-plate is the process of adjusting the pattern according to the different body types of customers. The process of scaling the pattern to make samples of various specifications is called clothing push-plate, also called clothing sizing.
[0003] At present, traditional clothing pushers have problems such as reliance on manual labor, long time consumption, large errors, and low efficiency of clothing customization. Summary of the invention
[0004] The purpose of this application is to provide an intelligent push plate method, device, medium and product for clothing customization to improve the efficiency of clothing customization
[0005] To achieve the above objectives, this application provides the following solutions:
[0006] In a first aspect, the present application provides an intelligent push plate method for clothing customization, comprising:
[0007] According to the net human dimensions of models of different body types, a paper pattern pushing method is adopted to construct size templates of models of different body types, and key points of the size templates of models of different body types are extracted; the size template includes multiple samples, each sample is composed of multiple coordinate points and structural lines; the key points are: the inflection point of the outer contour line, or the turning point between the arc and the straight line, or the intersection point of the line connecting the coordinate points and the structural line; the net human dimensions of models of different body types are used as the input of the neural network model, and the key points of the size templates of models of different body types are used as labels to train the neural network model to obtain a trained neural network model; according to the net human dimensions of the customer, the key points of the customer's size template are obtained by using the trained neural network model; according to the key points of the customer's size template, each key point in the reference size template is scaled to obtain the customer's customized clothing template; the version number of the reference size template is 160 / 84A.
[0008] In a second aspect, the present application provides a computer device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement any of the above-described smart push plate methods for clothing customization.
[0009] In a third aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any of the above-described intelligent push plate methods for clothing customization.
[0010] In a fourth aspect, the present application provides a computer program product, including a computer program, which, when executed by a processor, implements any of the above-mentioned smart push plate methods for clothing customization.
[0011] According to the specific embodiments provided in this application, this application discloses the following technical effects:
[0012] The present application obtains the net human dimensions of models with non-standard body shapes and key points in non-reference size samples, and uses the net human dimensions of models with non-standard body shapes as the input layer of a neural network model, and uses the key points in the non-reference size samples as the output layer of the neural network model, to train the neural network model to obtain a trained neural network model; the trained neural network model is used to customize clothing for customers to obtain customized clothing samples for customers, and only a few models need to be used for trial wearing to train the neural network model, so that fitting clothing can be customized for most customers. BRIEF DESCRIPTION OF THE DRAWINGS
[0013] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments 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 drawings can be obtained based on these drawings without paying creative work.
[0014] Figure 1 A schematic flow chart of an intelligent push plate method for clothing customization provided in an embodiment of the present application.
[0015] Figure 2 This is a diagram of the origin setting position of the template provided in the embodiment of the present application.
[0016] Figure 3 The coordinate points and key point distribution diagram of the samples in the template provided in the embodiment of the present application.
[0017] Figure 4 A flowchart for selecting key points is provided in an embodiment of the present application.
[0018] Figure 5 An embodiment of the present application provides a schematic diagram of serial number alignment between a reference size template and a non-reference size template.
[0019] Figure 6 A schematic diagram of the structure of a trained neural network model provided in an embodiment of the present application.
[0020] Figure 7 Schematic diagram of the paper pattern of the sleeve provided for the application embodiment.
[0021] Figure 8 Schematic diagram of the paper sample of the yoke provided for the application example.
[0022] Fig. 9 A schematic diagram of the structure of a computer device provided in an embodiment of the present application. DETAILED DESCRIPTION
[0023] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.
[0024] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.
[0025] At present, traditional clothing push-plates rely on manual work, which is time-consuming and has large errors. Computer-aided design of digital push-plates can solve some problems, but existing methods have shortcomings. Basically, they use mechanical processes such as formula methods and cannot work automatically and intelligently according to the customer's body shape.
[0026] like Figure 1 As shown, this embodiment provides an intelligent push plate method for clothing customization, including:
[0027] S1. Figure 2-Figure 5 As shown, according to the net body dimensions of models with different body shapes, a paper pattern pushing method is adopted to construct size templates for models with different body shapes, and key points of the size templates for models with different body shapes are extracted; the size template includes multiple samples, each sample is composed of multiple coordinate points and structural lines; the key points are: the inflection point of the outer contour line, or the turning point between the arc and the straight line, or the intersection of the line connecting the coordinate points and the structural line.
[0028] In actual application, Figure 2 As shown, all the lines in the size template are composed of points. For example, the contour line of the yoke in the figure is composed of dense points, rather than a description of a curve. Therefore, we need to find the red key points. The blue ones are coordinate points that follow the red changes.
[0029] Furthermore, the key points of the size samples of models with different body shapes are extracted, including:
[0030] Among the sizes of models with different body shapes, one is selected as a reference size sample, and the rest are selected as non-reference size samples.
[0031] Determine whether the serial numbers of the key points in the reference size template are consistent with those in the non-reference size template. If so, obtain the key points in the non-reference human body net size. If not, align the serial numbers of the key points in the non-reference size template with those in the non-reference size template to obtain the aligned key points in the non-reference size template.
[0032] Further, the key points in the non-reference size template are aligned with the key points in the non-reference size template in order to obtain the aligned key points in the non-reference size template, specifically including:
[0033] The serial numbers of the key points to be aligned in the non-reference size template are mapped to the serial numbers of the key points in the corresponding reference size template, and the coordinates of the key points to be aligned in the non-reference size template remain unchanged to obtain the key points in the aligned non-reference size template; the key points to be aligned are key points in the non-reference size template with different serial numbers from the key points in the reference size template.
[0034] S2. Figure 6 As shown, the net body dimensions of models of different body shapes are used as inputs of the neural network model, and the key points of the size samples of models of different body shapes are used as labels to train the neural network model to obtain a trained neural network model.
[0035] S3. According to the customer's net body size, use the trained neural network model to obtain the key points of the customer's size sample.
[0036] S4. According to the key points of the customer's size sample, each key point in the reference size sample is scaled to obtain the customer's customized clothing sample; the version number of the reference size sample is 160 / 84A.
[0037] Furthermore, the sample piece is a figure with a closed contour line.
[0038] Furthermore, the algorithm used in scaling the reference size template is an interpolation algorithm.
[0039] Furthermore, the net human body dimensions include at least: chest circumference, waist circumference, hip circumference, front length, back length, shoulder width, shoulder slope, neck circumference and arm length.
[0040] Furthermore, the dimension template is a DXF file.
[0041] Taking women's V-neck casual shirts and men's shirts as examples, the specific steps of the intelligent push plate method for clothing customization are as follows:
[0042] The first step is to create a specification table size_table.
[0043] A specification table is established based on the values of the 9 selected parts, namely: chest circumference, waist circumference, hip circumference, front length, back length, shoulder width, shoulder slope, neck circumference, and arm length.
[0044] Step 2: Make a reference size sample (standard paper pattern structure).
[0045] According to the industry standard 160 / 84A, the standard electronic template is drawn in the CAD software to form a DXF file of the reference size template.
[0046] Step 3: Read the DXF file of the reference size template and standardize the data.
[0047] In order to prevent the pattern coordinates from affecting and causing errors in subsequent calculations, normalization processing is required. The specific process is as follows: extract and sort out the key information of the elements in the DXF file of the reference size template. The key information specifically includes: the stored text in the DXF file of the reference size template and the point coordinates corresponding to the closed line segments on each sample. The reading process is:
[0048] 1. Check whether the figure in the reference size template is a closed figure. If so, it is determined to be a sample.
[0049] 2. Read the content on the sample, and read the specific text on the sample and the coordinates of the text placement.
[0050] 3. Read the line segments that make up the sample, divide them into multiple points, and output the corresponding coordinates of each coordinate point.
[0051] 4. Convert the pattern data after successful reading into JSON format.
[0052] The paper sample data in JSON format is as follows:
[0053] Taking the example of a women's shirt sleeve, the converted JSON file is:
[0054] (1)[{
[0055] (2) "type":"pattern",
[0056] (3) "name":"sleeves",
[0057] (4) "location":{0,0},
[0058] (5) "entity":[
[0059] (6) {
[0060] (7) "entity":"POLYLINE",
[0061] (8) "closed":true,
[0062] (9) "vertices":[[[305.6232910000001,5.030884000000015],[305.61908000000005,10.539672999999993,]......],
[0063] (10) "keypoints":[]
[0064] (11)},{ (12) ......
[0066] (13)} (14) ]
[0068] },{ (16)......
[0070] }]
[0071] Among them, the sleeves are patterned like Figure 7 shown.
[0072] For men's shirts, read the DXF file and save the pattern data in it in JSON format, such as: (1) [{
[0073] (2) "type":"pattern",
[0074] (3) "name":"Yuke",
[0075] (4) "location":{0,0},
[0076] (5) "entity":[
[0077] (6) {
[0078] (7) "entity":"POLYLINE",
[0079] (8) "closed":true,
[0080] (9) "vertices":[(236,65),(250,65),...,(222,65)],
[0081] (10) "keypoints":[]
[0082] (11)},{ (12) ......
[0084] (13)} (14) ]
[0086] },{ (16)......
[0088] }]
[0089] Among them, the pattern of the yoke is as follows Figure 7 shown.
[0090] 5. Take the coordinate point of the lower left corner of all samples in the reference size template as the origin.
[0091] 6. According to the difference between the coordinates of the key point in the lower left corner of all samples in the reference size template and the coordinates of the origin, the coordinates of all coordinate points are normalized, the coordinates of all coordinate points are updated, and the normalized coordinate points are obtained; the coordinates of the origin are (0, 0).
[0092] Step 4: Check the key points in the reference size template.
[0093] The "closed" property of the exported pattern outline is true, and that of other images is false, so that the true pattern outline point set can be found smoothly.
[0094] The method for finding key points in coordinate points is as follows:
[0095] 1. Set the coordinate point number i=1;
[0096] 2. For the i-th vertex in the coordinate point, get the coordinates and save them to (x1, y1);
[0097] 3. If i is not the last vertex, get the coordinates of the i+1th vertex and save them to (x2, y2); otherwise, save the coordinates of the first vertex to (x2, y2);
[0098] 4. Calculate the angle a of the straight line from (x1, y1) to (x2, y2);
[0099] 5. If the angle a is greater than 40°, set the i-th vertex as a key point and save the sequence number of i to keypoints.
[0100] 6.i=i+1, if i is greater than the total number of vertices, end; otherwise, go to step 2.
[0101] In this way, you can add the key point information in the pattern data, where the numbers are the sequence numbers of the vertices:
[0102] Men's shirt yoke example key points "keypoints": [0,26,28,38,40,50,52].
[0103] Example keypoints of a shirt sleeve: [0,58,109,174,232].
[0104] Step 5: Create size templates for 10-30 models with different body shapes (personalized fit).
[0105] Complete the reference size sample (standard sample), find 10-30 adult women and men of different body shapes, input their net body dimensions into the specification table established in the first step, make the garment, and try it on for fit. By constantly adjusting the pattern, determine the fit sample data, and form the sample data into a CAD file. Here, we still use the existing clothing CAD software to adjust the size of the clothing sample according to the different body shapes, that is, use the CAD software to call out the fit sample and export the file in DXF format.
[0106] Step 6: Align data for non-reference size templates (personalized templates).
[0107] The non-reference size samples of the fitting model are processed in the same way as the reference size samples to form a set of key points for each personalized sample. However, due to the deformation of the sample, the sequence number of the key point set obtained may be different from the key points of the standardized sample. For example, the following key point sequence numbers may be obtained: [0, 26, 27, 38, 42, 49, 54].
[0108] There will be some differences in the key points of the reference size sample. According to experiments, the sequence number will be different, but the number will not change. Therefore, data alignment is to check whether the number of key points detected is consistent with the number of key points of the standard sample. If they are consistent, only a simple mapping is needed. The mapping process is as follows Figure 5 shown.
[0109] Step 7: Model training.
[0110] In the previous section, we have formed a size_table of specifications for 10-30 models, which contains data of 9 dimensions and samples of different sizes. The sample data is processed in the same way as above to form a set of key points. The specification table and the set of key points form a training sample set to build a neural network model. After inputting the specification table data, we get the corresponding key point coordinates and finally get a trained neural network model.
[0111] Step 8. Use the difference to draw the template vertices.
[0112] After obtaining the prediction model of key points, the 9-dimensional body measurement data of each customer is used to form a specification table, and the coordinates of the key points of the sample are calculated using the prediction model. Then, using the interpolation principle, according to the vertex set of the standard sample, the displacement of the key points before and after is found, and the new coordinates of each vertex are obtained by the interpolation algorithm, and the customized clothing sample of the customer is finally obtained.
[0113] The present application obtains the net human dimensions of models with non-standard body shapes and key points in non-reference size samples, and uses the net human dimensions of models with non-standard body shapes as the input layer of a neural network model, and uses the key points in the non-reference size samples as the output layer of the neural network model, to train the neural network model to obtain a trained neural network model; the trained neural network model is used to customize clothing for customers to obtain customized clothing samples for customers, and only a few models need to be used for trial wearing to train the neural network model, so that fitting clothing can be customized for most customers.
[0114] Embodiment 2: The present application also provides a computer device, which may be a server or a terminal, and its internal structure diagram may be as follows: 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. Among them, 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. Among them, 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 video tag processing data. 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, the above-mentioned intelligent push plate method for clothing customization is implemented.
[0115] Those skilled in the art will understand that Fig. 9 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.
[0116] Embodiment 3, the present invention further provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the above-mentioned intelligent push plate method for clothing customization is implemented.
[0117] In a fourth embodiment, the present invention further provides a computer program product, including a computer program, which, when executed by a processor, implements the above-mentioned intelligent push plate method for clothing customization.
[0118] 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.
[0119] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods 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 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. By way of illustration and not limitation, RAM may be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).
[0120] The technical features of the above embodiments may be arbitrarily combined. 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 specification.
[0121] This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and core ideas of this application. At the same time, for those skilled in the art, according to the ideas of this application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.
Claims
1. An intelligent push plate method for clothing customization, characterized in that: The intelligent push plate method for clothing customization includes: According to the net body dimensions of models of different body types, a paper pattern pushing method is adopted to construct size templates of models of different body types, and key points of the size templates of models of different body types are extracted; the size template includes a plurality of samples, each sample is composed of a plurality of coordinate points and structural lines; the key points are: the inflection point of the outer contour line, or the turning point between the arc and the straight line, or the intersection point of the line connecting the coordinate points and the structural line; The net body dimensions of models of different body shapes are used as inputs of the neural network model, and the key points of the dimension samples of models of different body shapes are used as labels to train the neural network model to obtain a trained neural network model; According to the customer's body size, use the trained neural network model to obtain the key points of the customer's size sample; According to the key points of the customer's size sample, each key point in the reference size sample is scaled to obtain the customer's customized clothing sample; the version number of the reference size sample is 160 / 84A.
2. The intelligent push plate method for clothing customization according to claim 1 is characterized in that: Extract the key points of the size samples of models with different body shapes, including: Select one of the sizes of models with different body shapes as the reference size sample, and the rest as non-reference size samples; Determine whether the serial numbers of the key points in the reference size template are consistent with those in the non-reference size template. If so, obtain the key points in the non-reference human body net size. If not, align the serial numbers of the key points in the non-reference size template with those in the non-reference size template to obtain the aligned key points in the non-reference size template.
3. The intelligent push plate method for clothing customization according to claim 2 is characterized in that: Align the key points in the non-reference dimension template with the key points in the non-reference dimension template by serial number, and obtain the key points in the aligned non-reference dimension template, specifically including: The serial numbers of the key points to be aligned in the non-reference size template are mapped to the serial numbers of the key points in the corresponding reference size template, and the coordinates of the key points to be aligned in the non-reference size template remain unchanged to obtain the key points in the aligned non-reference size template; the key points to be aligned are key points in the non-reference size template with different serial numbers from the key points in the reference size template.
4. The intelligent push plate method for clothing customization according to claim 1, characterized in that: The sample piece is a figure with a closed contour line.
5. The intelligent push plate method for clothing customization according to claim 1, characterized in that: The algorithm used in scaling the reference dimension template is an interpolation algorithm.
6. The intelligent push plate method for clothing customization according to claim 1, characterized in that: The net human body measurements include at least: chest circumference, waist circumference, hip circumference, front length, back length, shoulder width, shoulder slope, neck circumference and arm length.
7. The intelligent push plate method for clothing customization according to claim 1, characterized in that: The dimension template is a DXF file.
8. A computer device comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the smart push plate method for clothing customization described in any one of claims 1-7.
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 intelligent push plate method for clothing customization described in any one of claims 1-7 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 intelligent push plate method for clothing customization described in any one of claims 1-7 are implemented.