Garment template generation method and system based on big data analysis
By using big data analytics, a clothing record database, and similarity weighting, the problem of clothing pattern design relying on experience was solved, resulting in more efficient and personalized clothing pattern generation.
Patent Information
- Application Number
- CN202511394242.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-09-28
- Publication Date
- 2026-01-23
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Existing garment pattern design relies heavily on the professional experience of senior pattern makers, resulting in high design thresholds, poor design effects, and excessive similarity.
Using a big data analysis approach, the design types of the garments are obtained, and the garments are sorted using evaluation parameters from the garment record database. The local features of the garments are broken down and randomly combined to form virtual garment patterns. Then, by judging similarity weights and evaluation parameters, designs with high similarity are eliminated, and effective garment patterns are output.
It lowers the barrier to entry for garment pattern design, reduces overly similar designs, and enables more efficient garment pattern generation.
Smart Images
Figure CN121389428A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of clothing design technology, and in particular to a method and system for generating clothing patterns based on big data analysis. Background Technology
[0002] Garment pattern design is a core part of the garment manufacturing process. The rationality and accuracy of its design directly determine the fit, comfort, aesthetics, and market acceptance of the final garment.
[0003] Currently, the mainstream garment pattern design methods in the industry heavily rely on the professional experience and subjective judgment of senior pattern makers. Specifically, existing techniques typically involve pattern makers creating patterns manually or with the assistance of CAD software, based on design renderings or style drawings, combined with their personal experience, understanding of human anatomy, mastery of fabric characteristics, and perception of fashion trends.
[0004] The current pattern-making method requires experienced staff, resulting in a high barrier to entry for garment pattern design. Furthermore, individual designers' subjective design habits can lead to numerous similar design elements across different garments, resulting in poor design outcomes. Therefore, there is an urgent need to design a method that facilitates garment pattern design to achieve effective garment pattern creation. Summary of the Invention
[0005] To achieve effective design of garment patterns, this application provides a method and system for generating garment patterns based on big data analysis.
[0006] Firstly, this application provides a method for generating garment patterns based on big data analysis, employing the following technical solution: A method for generating garment patterns based on big data analysis, comprising: Obtain the design type of the requirements; Based on the design requirements, the existing garment patterns and garment evaluation parameters are determined from the pre-set garment record library; The existing garment patterns are sorted according to the garment evaluation parameters from largest to smallest to determine the garment type ranking, and the existing garment patterns that rank higher in the garment type ranking are defined as high-quality garment patterns based on the preset reference quantity. Based on the preset part disassembly rules, the local features of each high-quality pattern garment are determined, and the local features of the same part are combined to construct an effective feature set; Randomly select a local feature of clothing from each set of valid features as the local feature to be used, and combine all the local features to determine the virtual pattern clothing according to the part decomposition rules; The virtual pattern garments are analyzed to determine the virtual evaluation parameters. Virtual pattern garments with virtual evaluation parameters greater than the preset benchmark evaluation parameters are defined as valid pattern garments and the valid pattern garments are output.
[0007] Optionally, after the virtual garment pattern is determined, the garment pattern generation method based on big data analysis also includes: The virtual pattern garment is compared with a random single existing pattern garment to define the same local features of the garment as common local features; The local feature type is determined based on common local features, and the local similarity weight corresponding to the local feature type is determined based on the preset similarity matching relationship. The overall similarity weight is determined by calculating all local similarity weights, and it is then determined whether the overall similarity weight is greater than the preset viewing similarity weight. If the overall similarity weight is not greater than the viewing similarity weight, then the currently determined virtual clothing pattern will be maintained. If the overall similarity weight is greater than the viewing similarity weight, the current virtual clothing pattern will be removed.
[0008] Optionally, the step of calculating the overall similarity weight based on all local similarity weights includes: A common local feature is randomly selected as the principal local feature, and the common local features adjacent to the principal local feature are updated as the principal local features. The local cluster combination is determined by induction based on all the principal local features. The number of clusters within a single local cluster combination is determined by counting based on common local characteristics. The weight compensation multiple corresponding to the number of items within the cluster is determined based on the preset compensation matching relationship. The local similarity weights are updated by calculating the local similarity weights based on the common local features within the local cluster combination and the corresponding weight compensation multiples. The overall similarity weight is then determined by summing the updated local similarity weights.
[0009] Optionally, the steps of analyzing virtual garment patterns to determine virtual evaluation parameters include: In the virtual pattern of clothing, random combinations are made based on the local features of the clothing to construct matching feature combinations, and the number of items in the matching feature combinations is determined by counting the local features of the clothing within the matching feature combinations. In the existing patterned clothing, existing patterned clothing containing matching feature combinations is defined as historical matching clothing, and the number of historical matches is counted based on the historical matching clothing, and the proportion of historical matches is determined based on the number of historical matches and the preset number of registrations. The effective pairing weights corresponding to the number of pairs within a pairing are determined based on the preset weight matching relationship, and the reasonable coefficient of a single pairing is calculated based on the effective pairing weights and the corresponding historical pairing ratios. The virtual evaluation parameters are determined by summing up the reasonable coefficients of all individual components.
[0010] Optionally, the step of randomly combining local features of the virtual garment pattern to construct a combination of matching features includes: Randomly select one local feature from all local features of clothing as the starting local feature, and define the corresponding local similarity weight as the starting similarity weight according to the local feature type of the starting local feature. The remaining similarity weight is determined by calculating the difference between the viewing similarity weight and the initial similarity weight. Among the clothing local features that are not the starting local features, clothing local features with local similarity weights less than the remaining similarity weights are combined to determine optional feature combinations. Then, a clothing local feature is randomly selected from the optional feature combinations and added to the starting local features to adjust the remaining similarity weights. This process continues until the optional feature combinations are empty, at which point a matching feature combination is constructed based on all the starting local features.
[0011] Optionally, after the number of historical matches is determined, the method for generating clothing patterns based on big data analysis also includes: The clothing evaluation parameters corresponding to historically matched clothing are defined as historical evaluation parameters; The matching importance coefficient corresponding to the historical evaluation parameters is determined based on the preset importance matching relationship; The historical match count is updated based on all matching importance coefficients and the number of historical matches.
[0012] Secondly, this application provides a garment pattern generation system based on big data analysis, employing the following technical solution: A garment pattern generation system based on big data analysis includes: The acquisition module is used to acquire the design type of the requirements; The processing module, connected to the acquisition and judgment modules, is used for information storage and processing; The judgment module, connected to the acquisition and processing modules, is used for judging information. The processing module determines the existing garment patterns and garment evaluation parameters from the preset garment record library based on the required design type. The processing module sorts the existing garment patterns according to the garment evaluation parameters from largest to smallest to determine the garment type ranking, and defines the existing garment patterns that rank higher in the garment type ranking as high-quality garment patterns based on the preset reference quantity. The processing module determines the local features of each high-quality garment pattern according to the preset part disassembly rules, and combines the local features of the same part to construct an effective feature set; The processing module randomly selects a local feature of clothing from each set of valid features as the local feature to be used, and combines all the local features to be used according to the part decomposition rules to determine the virtual pattern clothing; The processing module analyzes the virtual pattern garment to determine the virtual evaluation parameters, and defines the virtual pattern garments whose virtual evaluation parameters determined by the judgment module are greater than the preset benchmark evaluation parameters as valid pattern garments, and outputs the valid pattern garments.
[0013] Thirdly, this application provides a computer storage medium capable of storing corresponding programs, which has the characteristic of effectively designing garment patterns, and adopts the following technical solution: A computer-readable storage medium storing a computer program that can be loaded by a processor and executed any of the above-described methods for generating clothing patterns based on big data analysis.
[0014] In summary, this application includes at least one of the following beneficial technical effects: In the process of designing garment patterns, the market can be analyzed to identify similar garments with high ratings, and the layout can be automatically designed to create garments that may meet production requirements. This lowers the design threshold and enables effective garment pattern design. By analyzing and judging local similarities, the generation of excessively similar clothing samples can be reduced, thus achieving effective design of clothing patterns. Attached Figure Description
[0015] Figure 1 This is a flowchart of a method for generating clothing patterns based on big data analysis.
[0016] Figure 2 This is a flowchart of the module for a clothing pattern generation method based on big data analysis. Detailed Implementation
[0017] To make the purpose, technical solution, and advantages of this application clearer, the following is combined with Figures 1-2 The present application will be further described in detail below with reference to embodiments. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the application.
[0018] The embodiments of this application will now be described in further detail with reference to the accompanying drawings.
[0019] This application discloses a method for generating clothing patterns based on big data analysis, referring to... Figure 1The method for generating garment patterns based on big data analysis includes the following steps: Step S100: Obtain the requirement design type.
[0020] The design requirement type refers to the type of clothing that needs to be designed, such as short-sleeved shirts, long-sleeved shirts, windbreakers, etc.
[0021] Step S101: Based on the required design type, determine the existing pattern garments and garment evaluation parameters from the preset garment record library.
[0022] The clothing record database is a database that registers and records clothing sold in the market. Existing pattern clothing refers to clothing of the demand design type in the clothing record database. The clothing evaluation parameter is a parameter value that reflects the positive feedback of the existing pattern clothing. The higher the value, the more the existing pattern clothing meets market requirements. The clothing evaluation parameter can be obtained by collecting content from market evaluations, such as the positive feedback rate. The specific parameters can be set by staff according to the actual situation, and the corresponding data will be recorded in the clothing record database.
[0023] Step S102: Sort each existing pattern garment according to the garment evaluation parameters from largest to smallest to determine the garment type ranking, and define the existing pattern garments that rank higher in the garment type ranking as high-quality pattern garments according to the preset reference quantity.
[0024] The garment type sorting is the order of the relevant garments obtained after sorting the existing garment patterns. The sorting rule is that garments with larger evaluation parameters are ranked first, and garments with smaller evaluation parameters are ranked last. The reference quantity is a fixed number set by the staff. By defining high-quality garment patterns, the existing garment patterns with better evaluation in the current garment type are identified and distinguished to facilitate subsequent analysis.
[0025] Step S103: Determine the local features of each high-quality garment pattern according to the preset part disassembly rules, and combine the local features of the same part to construct an effective feature set.
[0026] Part decomposition rules refer to the methods of decomposing the parts contained in the current design type of clothing, such as sleeves, waist, collar, etc., so that a complete garment can be formed by integrating the various parts; local features of clothing are the part features obtained after high-quality patterned clothing is decomposed according to the part decomposition rules; effective feature set is the set formed by combining local features of the same part of clothing.
[0027] Step S104: Randomly select a local feature of clothing from each set of valid features as the local feature to be used, and combine all the local features to be used according to the part decomposition rules to determine the virtual pattern clothing.
[0028] By randomly defining and integrating local features, a complete garment in theoretical terms can be obtained, i.e., a virtual garment pattern. By randomly selecting and using local features, various garments that meet the requirements can be designed.
[0029] Step S105: Analyze the virtual pattern garment to determine the virtual evaluation parameters, define the virtual pattern garments whose virtual evaluation parameters are greater than the preset benchmark evaluation parameters as valid pattern garments, and output the valid pattern garments.
[0030] Virtual evaluation parameters are parameter values that reflect the reliability or reasonableness of the currently generated virtual pattern garment. They can be manually entered by staff or determined through steps S400-S403. The benchmark evaluation parameter is the minimum virtual evaluation parameter that the staff sets to determine when the overall garment is considered relatively reliable and reasonable. Therefore, defining effective pattern garments distinguishes virtual pattern garments that meet the requirements, thereby enabling the output of effective pattern garments and realizing the effective design of garment patterns.
[0031] After the virtual garment pattern is determined, the garment pattern generation method based on big data analysis also includes: Step S200: Compare the virtual pattern garment with a random single existing pattern garment to define the same local features of the garment as common local features.
[0032] Common local features are defined to identify the same local features of two garments when comparing them, which facilitates subsequent analysis.
[0033] Step S201: Determine the local feature type based on common local features, and determine the local similarity weight corresponding to the local feature type based on the preset similarity matching relationship.
[0034] Local feature types are part types that share common local features, such as sleeves. Different parts have different visual impacts on users due to their different areas, and the corresponding coefficients reflecting the similarity between two garments are also different. Local similarity weight is the coefficient that reflects the similarity between two garments. Different local feature types correspond to different local similarity weights. The similarity matching relationship between the two is determined in advance by the staff and recorded and stored.
[0035] Step S202: Calculate the overall similarity weight based on all local similarity weights, and determine whether the overall similarity weight is greater than the preset viewing similarity weight.
[0036] The overall similarity weight is a parameter value that reflects the overall similarity between two garments. The larger the value, the more similar the two garments are. The visual similarity weight is the minimum overall similarity weight that the staff sets when they consider two garments to be visually similar. The purpose of the judgment is to determine whether the current virtual garment pattern is similar to an existing garment, and thus determine whether the garment should be processed into a template.
[0037] Step S2021: If the overall similarity weight is not greater than the viewing similarity weight, then maintain the currently determined virtual clothing pattern.
[0038] When the overall similarity weight is not greater than the viewing similarity weight, it indicates that there is a certain difference between the current virtual clothing pattern and the existing clothing. Therefore, we can maintain the currently determined virtual clothing pattern and continue the analysis.
[0039] Step S2022: If the overall similarity weight is greater than the viewing similarity weight, then the current virtual clothing pattern will be removed.
[0040] When the overall similarity weight is greater than the viewing similarity weight, it means that the current virtual clothing pattern is quite similar to the existing clothing and there is no need for layout. Therefore, it is removed to reduce the amount of invalid data analysis.
[0041] The steps for determining the overall similarity weight based on all local similarity weights include: Step S300: Randomly select a common local feature as the main local feature, update the common local features adjacent to the main local feature as the main local feature, and summarize based on all the main local features to determine the local cluster combination.
[0042] By defining key local features to determine the common local features of connected elements, local cluster combinations are constructed to determine the common local features in each phase set, which facilitates subsequent analysis. For example, if there are common local features A, B, C, and D, where B is adjacent to A and D respectively, and C is not adjacent to any of A, B, or D, then there will be two local cluster combinations, namely ABD and C.
[0043] Step S301: Count the number of clusters within a single local cluster combination based on common local features.
[0044] The number of clusters within a cluster refers to the number of common local features present in a combination of local clusters. Taking the example above, the number of clusters within a cluster corresponding to the local cluster combination of ABD is 3.
[0045] Step S302: Determine the weight compensation multiple corresponding to the number of items within the cluster based on the preset compensation matching relationship.
[0046] The weight compensation factor is a parameter value used to calculate and compensate for the local similarity weight. The larger the number of clusters, the more connected the parts are, resulting in a greater visual impact. Compared to the separate features, they are more similar, so the local similarity weight needs to be compensated. The larger the number of clusters, the larger the corresponding weight compensation factor. The compensation matching relationship between the two is determined in advance by the staff and will not be elaborated here.
[0047] Step S303: Calculate and update the local similarity weights based on the local similarity weights of the common local features within the local cluster combination and the corresponding weight compensation multiples, and sum the updated local similarity weights to determine the overall similarity weight.
[0048] The local similarity weights can be updated by multiplying the local similarity weights of common local features by the corresponding weight compensation factor. Then, by summing all the updated local similarity weights, we can obtain the overall similarity weight that more accurately reflects the similarity situation.
[0049] The steps for determining virtual evaluation parameters based on virtual garment patterns include: Step S400: In the virtual pattern of clothing, random combinations are made based on the local features of the clothing to construct a matching feature combination, and the number of items in the matching feature combination is determined by counting the local features of the clothing within the matching feature combination.
[0050] The combination of features is a feature combination formed by randomly selecting at least two local features of clothing from the virtual pattern clothing. The number of features in the combination is the number of local features of clothing contained in the combination of features.
[0051] Step S401: Define existing patterned garments containing matching feature combinations as historical matching garments, count the number of historical matches based on historical matching garments, and calculate the historical matching percentage based on the number of historical matches and the preset number of registrations.
[0052] Historical matching garments are defined to distinguish different existing garment patterns. The number of historical matches is the number of garments with the current matching feature combination, which can be determined by counting each historical matching garment. The number of registrations is the total number of existing garment patterns identified in the garment record database. The historical matching percentage is the proportion of historical matching garments to existing garment patterns, which can be determined by dividing the number of historical matches by the number of registrations.
[0053] Step S402: Determine the effective matching weights corresponding to the number of items in the matching based on the preset weight matching relationship, and calculate the reasonable coefficient of the individual matching based on the effective matching weights and the corresponding historical matching proportions.
[0054] Effective combination weight is a parameter value that reflects the reasonable relationship of combinations. The smaller the number of combinations, the smaller the reasonable situation, and the smaller the corresponding effective combination weight. The weight matching relationship between the two is determined in advance by the staff. The individual combination reasonableness coefficient is obtained by multiplying the effective combination weight by the corresponding historical matching ratio, and then multiplying by the preset adjustment coefficient that reflects the importance of the historical matching ratio. That is, the parameter value that reflects the reasonableness of the current combination features when they are combined together.
[0055] Step S403: Summate and calculate the virtual evaluation parameters based on the reasonable coefficients of all individual components.
[0056] By adding up the reasonableness coefficients of all individual components, we can obtain a parameter value that reflects whether the current clothing is reliable and reasonable, i.e., a virtual evaluation parameter.
[0057] The steps for constructing matching feature combinations in virtual garment patterns by randomly combining local features of the garment include: Step S500: Randomly select a local feature of clothing from all local features of clothing as the starting local feature, and define the corresponding local similarity weight as the starting similarity weight according to the local feature type of the starting local feature.
[0058] Define initial local features and initial similarity weights to distinguish different data, which facilitates subsequent analysis.
[0059] Step S501: Calculate the difference based on the viewing similarity weight and the initial similarity weight to determine the remaining similarity weight.
[0060] The remaining similarity weight is the value obtained by subtracting the initial similarity weight from the viewing similarity weight.
[0061] Step S502: Among the clothing local features that are not the starting local features, combine the clothing local features whose local similarity weight is less than the remaining similarity weight to determine the optional feature combination. Then, randomly select a clothing local feature from the optional feature combination and add it to the starting local feature to adjust the remaining similarity weight. Continue until the optional feature combination is empty, and then construct the matching feature combination based on all the starting local features.
[0062] By determining the optional feature combinations, we can identify the clothing local features that can be combined with the current starting local features to construct a matching feature combination. This allows for feature selection within the optional range, reducing the occurrence of invalid data analysis. By continuously updating the remaining similarity weights, we can simulate the continuous selection of clothing local features. When the optional feature combinations are empty, it means that no more clothing local features can be added. At this time, the determined starting local features can be combined to construct the required matching feature combination.
[0063] After determining the number of historical matches, the clothing pattern generation method based on big data analysis also includes: Step S600: Define the clothing evaluation parameters corresponding to the historically matched clothing as historical evaluation parameters.
[0064] Define historical evaluation parameters to differentiate between different data points, facilitating subsequent analysis.
[0065] Step S601: Determine the matching importance coefficient corresponding to the historical evaluation parameters based on the preset importance matching relationship.
[0066] The matching importance coefficient is a coefficient that reflects the influence of different clothing evaluations on the reliability and rationality of virtual pattern clothing. The larger the historical evaluation parameter, the larger the corresponding matching importance coefficient. The importance matching relationship between the two is determined in advance by the staff and entered into the storage.
[0067] Step S602: Calculate and update the historical match count based on all matching importance coefficients and the number of historical matches.
[0068] By averaging all the matching importance coefficients and then multiplying the average by the historical matching count, a more accurate historical matching count can be obtained, thereby improving the accuracy of data analysis.
[0069] Reference Figure 2 Based on the same inventive concept, embodiments of the present invention provide a clothing pattern generation system based on big data analysis, comprising: The acquisition module is used to acquire the design type of the requirements; The processing module, connected to the acquisition and judgment modules, is used for information storage and processing; The judgment module, connected to the acquisition and processing modules, is used for judging information. The processing module determines the existing garment patterns and garment evaluation parameters from the preset garment record library based on the required design type. The processing module sorts the existing garment patterns according to the garment evaluation parameters from largest to smallest to determine the garment type ranking, and defines the existing garment patterns that rank higher in the garment type ranking as high-quality garment patterns based on the preset reference quantity. The processing module determines the local features of each high-quality garment pattern according to the preset part disassembly rules, and combines the local features of the same part to construct an effective feature set; The processing module randomly selects a local feature of clothing from each set of valid features as the local feature to be used, and combines all the local features to be used according to the part decomposition rules to determine the virtual pattern clothing; The processing module analyzes the virtual pattern clothing to determine the virtual evaluation parameters, and defines the virtual pattern clothing whose virtual evaluation parameters determined by the judgment module are greater than the preset benchmark evaluation parameters as valid pattern clothing, and outputs the valid pattern clothing. The similar clothing removal module is used to remove clothing with a high degree of similarity, so as to achieve effective design of clothing patterns; The overall similarity weight calculation module is used to determine the overall similarity weight between garments; The virtual evaluation parameter determination module is used to determine the virtual evaluation parameters for each generated virtual garment pattern. The feature combination construction module is used to effectively construct feature combinations, reducing the amount of data analysis. The historical match count update module updates the historical match count based on the evaluation of each garment.
[0070] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional modules is used as an example. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. The specific working process of the system, device, and unit described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.
[0071] This invention provides a computer-readable storage medium storing a computer program that can be loaded by a processor and executed as a method for generating clothing patterns based on big data analysis.
[0072] Computer storage media include, for example, USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, optical disks, and other media that can store program code.
Claims
1. A method for generating garment patterns based on big data analysis, characterized in that, include: Obtain the design type of the requirements; Based on the design requirements, the existing garment patterns and garment evaluation parameters are determined from the pre-set garment record library; The existing garment patterns are sorted according to the garment evaluation parameters from largest to smallest to determine the garment type ranking, and the existing garment patterns that rank higher in the garment type ranking are defined as high-quality garment patterns based on the preset reference quantity. Based on the preset part disassembly rules, the local features of each high-quality pattern garment are determined, and the local features of the same part are combined to construct an effective feature set; Randomly select a local feature of clothing from each set of valid features as the local feature to be used, and combine all the local features to determine the virtual pattern clothing according to the part decomposition rules; The virtual pattern garments are analyzed to determine the virtual evaluation parameters. Virtual pattern garments with virtual evaluation parameters greater than the preset benchmark evaluation parameters are defined as valid pattern garments and the valid pattern garments are output.
2. The method for generating garment patterns based on big data analysis according to claim 1, characterized in that, After the virtual garment pattern is determined, the garment pattern generation method based on big data analysis also includes: The virtual pattern garment is compared with a random single existing pattern garment to define the same local features of the garment as common local features; The local feature type is determined based on common local features, and the local similarity weight corresponding to the local feature type is determined based on the preset similarity matching relationship. The overall similarity weight is determined by calculating all local similarity weights, and it is then determined whether the overall similarity weight is greater than the preset viewing similarity weight. If the overall similarity weight is not greater than the viewing similarity weight, then the currently determined virtual clothing pattern will be maintained. If the overall similarity weight is greater than the viewing similarity weight, the current virtual clothing pattern will be removed.
3. The method for generating garment patterns based on big data analysis according to claim 2, characterized in that, The steps to determine the overall similarity weight based on all local similarity weights include: A common local feature is randomly selected as the principal local feature, and the common local features adjacent to the principal local feature are updated as the principal local features. The local cluster combination is determined by induction based on all the principal local features. The number of clusters within a single local cluster combination is determined by counting based on common local characteristics. The weight compensation multiple corresponding to the number of items within the cluster is determined based on the preset compensation matching relationship. The local similarity weights are updated by calculating the local similarity weights based on the common local features within the local cluster combination and the corresponding weight compensation multiples. The overall similarity weight is then determined by summing the updated local similarity weights.
4. The method for generating garment patterns based on big data analysis according to claim 2, characterized in that, The steps for determining virtual evaluation parameters based on virtual garment patterns include: In the virtual pattern of clothing, random combinations are made based on the local features of the clothing to construct matching feature combinations, and the number of items in the matching feature combinations is determined by counting the local features of the clothing within the matching feature combinations. In the existing patterned clothing, existing patterned clothing containing matching feature combinations is defined as historical matching clothing, and the number of historical matches is counted based on the historical matching clothing, and the proportion of historical matches is determined based on the number of historical matches and the preset number of registrations. The effective matching weights corresponding to the number of items in a matching set are determined based on the preset weight matching relationship. The reasonable coefficient of a single matching set is then calculated based on the effective matching weights and the corresponding historical matching ratios. The virtual evaluation parameters are determined by summing up the reasonable coefficients of all individual components.
5. The method for generating garment patterns based on big data analysis according to claim 4, characterized in that, The steps for constructing matching feature combinations in virtual garment patterns by randomly combining local features of the garment include: Randomly select one local feature from all local features of clothing as the starting local feature, and define the corresponding local similarity weight as the starting similarity weight according to the local feature type of the starting local feature; The remaining similarity weight is determined by calculating the difference between the viewing similarity weight and the initial similarity weight. Among the clothing local features that are not the starting local features, clothing local features with local similarity weights less than the remaining similarity weights are combined to determine optional feature combinations. Then, a clothing local feature is randomly selected from the optional feature combinations and added to the starting local features to adjust the remaining similarity weights. This process continues until the optional feature combinations are empty, at which point a matching feature combination is constructed based on all the starting local features.
6. The method for generating garment patterns based on big data analysis according to claim 4, characterized in that, After determining the number of historical matches, the clothing pattern generation method based on big data analysis also includes: The clothing evaluation parameters corresponding to historically matched clothing are defined as historical evaluation parameters; The matching importance coefficient corresponding to the historical evaluation parameters is determined based on the preset importance matching relationship; The historical match count is updated based on all matching importance coefficients and the number of historical matches.
7. A garment pattern generation system based on big data analysis, characterized in that, include: The acquisition module is used to acquire the design type of the requirements; The processing module, connected to the acquisition and judgment modules, is used for information storage and processing; The judgment module, connected to the acquisition and processing modules, is used for judging information. The processing module determines the existing garment patterns and garment evaluation parameters from the preset garment record library based on the required design type. The processing module sorts the existing garment patterns according to the garment evaluation parameters from largest to smallest to determine the garment type ranking, and defines the existing garment patterns that rank higher in the garment type ranking as high-quality garment patterns based on the preset reference quantity. The processing module determines the local features of each high-quality garment pattern according to the preset part disassembly rules, and combines the local features of the same part to construct an effective feature set; The processing module randomly selects a local feature of clothing from each set of valid features as the local feature to be used, and combines all the local features to be used according to the part decomposition rules to determine the virtual pattern clothing; The processing module analyzes the virtual pattern garment to determine the virtual evaluation parameters, and defines the virtual pattern garments whose virtual evaluation parameters determined by the judgment module are greater than the preset benchmark evaluation parameters as valid pattern garments, and outputs the valid pattern garments.
8. A computer-readable storage medium, characterized in that, The computer program is stored and can be loaded by a processor and executed as described in any one of claims 1 to 6 for the method of generating clothing patterns based on big data analysis.