A human-clothing spatial data conversion method based on multi-feature cross-domain matching

Abstract

The present invention relates to the technical field of electronic digital data processing, and in particular to a method for converting spatial data of human clothing with multi-feature cross-domain matching. The method comprises obtaining measurement data, ages and style preferences of a plurality of individuals, and dividing the individuals into a plurality of groups according to their ages, wherein the measurement data comprises a plurality of body index data; determining the part significance of each body index data of each individual according to the measurement data, the style preferences and the total number of individuals in all groups, determining a target weight value of each body index data according to the part significance, and determining a target linear regression equation according to the target weight value; quantifying a corresponding target weight value based on the part significance of each body index data, so that when the target linear regression equation performs calculation analysis on the measurement data to be analyzed, the target parameters for designing clothing can be determined more accurately, which is conducive to meeting the size requirements of clothing for people of different body shapes.

Description

Technical Field

[0001] The present invention relates to the technical field of electronic digital data processing, and in particular to a method for converting human-clothing spatial data using multi-feature cross-domain matching. Background Art

[0002] With the development of science and technology, digital fashion design methods based on digital technology and intelligence have been widely used in the fashion design and manufacturing process. This design method combines the aesthetics and innovation of traditional fashion design, aiming to improve design efficiency, reduce costs, improve product quality, and meet the needs of personalized customization.

[0003] Usually in the digital fashion design method, the multi-feature cross-domain matching human-clothing spatial data conversion refers to the process of automatic conversion from the actual human body measurement data to clothing sample data. In this process, the human body measurement data of the survey needs to be analyzed by linear regression to determine the target parameters of the main parts of the clothing. For example, clothing can include dresses and pants such as jeans. However, due to the different body shapes and body proportions of different human bodies during the survey process, the body shapes of the surveyed people vary greatly. If the weights of the indicator data of different individuals are the same, the linear regression model will not be able to accurately predict the demand for clothing sizes for people of different body shapes. Summary of the invention

[0004] In order to solve the above technical problems, the purpose of the present invention is to provide a method for converting human-clothing spatial data into multi-feature cross-domain matching. The technical solution adopted is as follows:

[0005] In a first aspect, an embodiment of the present application provides a method for human-clothing spatial data conversion with multi-feature cross-domain matching, comprising:

[0006] Acquire measurement data, ages, and style preferences of a plurality of individuals, and divide the individuals into a plurality of groups according to the ages, wherein the measurement data includes a plurality of body index data;

[0007] Determine the site significance of each of the body index data of each of the individuals according to the measurement data, the style preference and the total number of individuals in all the groups;

[0008] Determine a target weight value of each of the body index data according to the site significance, and determine a target linear regression equation according to the target weight value;

[0009] The target parameters for designing the clothing are determined according to the measured data to be analyzed and the target linear regression equation.

[0010] In one embodiment, determining the site significance of each of the body indicator data of each of the individuals based on the measurement data, the style preference, and the total number of individuals in all of the groups includes:

[0011] Determining the group consistency of each individual in the group based on the measurement data and the pattern preference;

[0012] The site significance of each of the body indicator data of each of the individuals is determined based on the group consistency and the total number of individuals in all of the groups.

[0013] In one embodiment, determining the group consistency of each individual in the group according to the measurement data and the pattern preference includes:

[0014] Determining the consistency of characteristics between individuals in each of the groups based on the measurement data and the pattern preferences;

[0015] Using the feature consistency as a distance indicator, clustering each of the groups respectively to obtain a first subcluster and a second subcluster corresponding to each of the groups;

[0016] The group consistency of the individuals in each of the groups is determined according to the first sub-cluster, the second sub-cluster, the number of individuals in each of the groups, and the feature consistency.

[0017] In one embodiment, determining the consistency of characteristics between individuals in each group based on the measurement data and the pattern preference includes:

[0018] Calculate the characteristic similarity values ​​of the pattern preferences between individuals in each group according to the Jaccard coefficient and the pattern preferences corresponding to the individuals;

[0019] Calculate the absolute value of each corresponding difference in the physical indicator data between individuals in each group, and calculate the mean value of the difference according to the number of the physical indicator data and each absolute value of the difference;

[0020] A first calculation result is obtained according to the exponential function and the mean difference value, and the feature consistency between the individuals in each of the groups is obtained according to the first calculation result and the first product of the feature similarity value.

[0021] In one embodiment, determining the group consistency of individuals in each of the groups according to the first sub-cluster, the second sub-cluster, the number of individuals in each of the groups, and the feature consistency includes:

[0022] If the number of individuals in the first subcluster is greater than the number of individuals in the second subcluster, the first subcluster is used as the first target subcluster and the second subcluster is used as the second target subcluster; otherwise, the first subcluster is used as the second target subcluster and the second subcluster is used as the first target subcluster;

[0023] respectively determining a first ratio of the number of individuals in each of the first target sub-clusters to the number of individuals in the corresponding group;

[0024] Determining target feature consistency between individuals in the first target subcluster from the feature consistency, and determining a consistency mean corresponding to the target feature consistency;

[0025] A second calculation result is determined according to the normalization function and the consistency mean, and the group consistency of each individual in the group is determined according to a second product of the second calculation result and the first ratio.

[0026] In one embodiment, determining the site significance of each of the body indicator data of each of the individuals according to the group consistency and the total number of individuals in all the groups includes:

[0027] Determine the significance of the indicator of the individuals in each of the groups according to the total number of individuals in all of the groups, the number of individuals in each of the groups, the group consistency of the individuals in each of the groups, and the sum of the consistency of the group consistency of the individuals in all of the groups;

[0028] The site significance of each of the body indicator data of each of the individuals is determined according to the indicator significance and the body indicator data.

[0029] In one embodiment, determining the significance of the indicator of the individuals in each of the groups according to the total number of individuals in all of the groups, the number of individuals in each of the groups, the group consistency of the individuals in each of the groups, and the sum of the consistency of the group consistency of the individuals in all of the groups comprises:

[0030] Determining a second ratio of the group consistency of each individual of the group to the sum of the consistency of the group consistency of the individuals of all groups;

[0031] Determining a third ratio of the number of individuals in each of the groups to the total number of individuals in all of the groups;

[0032] The significance of the indicator of each individual in the group is determined according to a second product of the second ratio and the third ratio.

[0033] In one embodiment, determining the site significance of each of the body indicator data of each of the individuals according to the indicator significance and the body indicator data comprises:

[0034] Calculating the variance of each of the body index data respectively, and sorting the body index data from small to large according to the size of the variance;

[0035] When the body index data is longitudinal data along the long axis of the lower body of the human body, determine the sum of the first sequence numbers of the sorting sequence numbers of all the longitudinal data and the sorting sequence number of each longitudinal data, respectively determine the fourth ratio of the sorting sequence number of each longitudinal data to the sum of the first sequence numbers, respectively calculate the third product of the fourth ratio and the significance of the index corresponding to the individual, and obtain the first part significance of each longitudinal data of each individual;

[0036] When the body indicator data is transverse data of the dimensions on the cross-section of the lower body of the human body, determine the sum of the second serial numbers of the sorting numbers of all the transverse data and the sorting number of each of the transverse data, respectively determine the fifth ratio of the sorting number of each of the transverse data to the sum of the second serial numbers, respectively calculate the fourth product of the fifth ratio and the significance of the indicator corresponding to the individual, and obtain the second part significance of each transverse data of each of the individuals.

[0037] In one embodiment, determining a target weight value of each of the body index data according to the site significance, and determining a target linear regression equation according to the target weight value comprises:

[0038] Determine the sum of the first-part significances according to the total number of individuals in the group and the first-part significances of each longitudinal data of each individual, and obtain the target first weight value of each longitudinal data according to the sixth ratio of the sum of the first-part significances to the total number of individuals in the group;

[0039] Determine the sum of the second part significance according to the total number of individuals in the group and the second part significance of each horizontal data of each individual, and obtain the target second weight value of each horizontal data according to the seventh ratio of the sum of the second part significance to the total number of individuals in the group;

[0040] According to the target first weight value and the preset linear regression equation, a target first linear regression equation is obtained, and according to the target second weight value and the preset linear regression equation, a target second linear regression equation is obtained.

[0041] In one embodiment, determining the target parameters for designing clothing according to the measurement data to be analyzed and the target linear regression equation includes:

[0042] Determining a height value for designing clothing according to the longitudinal data to be analyzed in the measurement data to be analyzed and the target first linear regression equation;

[0043] Determining a hip circumference value of the designed clothing according to the to-be-analyzed transverse data in the to-be-analyzed measurement data and the target second linear regression equation;

[0044] The target parameters include height and hip circumference.

[0045] The present invention has the following beneficial effects:

[0046] The measurement data, age and style preference of several individuals are obtained, and the individuals are divided into multiple groups according to age, wherein the measurement data includes multiple body index data; the site significance of each body index data of each individual is determined according to the measurement data, style preference and the total number of individuals in all groups, and the target weight value of each body index data is determined according to the site significance, and the target linear regression equation is determined according to the target weight value, and the corresponding target weight value is quantified based on the site significance of each body index data, so that when the target linear regression equation performs calculation and analysis on the measurement data to be analyzed, the target parameters for designing clothing can be determined more accurately, which is conducive to meeting the size requirements of clothing for people of different body shapes. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0048] Figure 1 A schematic flowchart of the steps of a method for converting human-clothing spatial data by multi-feature cross-domain matching provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0049] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the specific implementation method, structure, features and effects of a multi-feature cross-domain matching human-clothing spatial data conversion method proposed by the present invention in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0050] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0051] The specific scheme of the human-clothing spatial data conversion method with multi-feature cross-domain matching provided by the present invention is described in detail below with reference to the accompanying drawings.

[0052] See also Figure 1 , which shows a flow chart of a method for converting human-clothing space data by multi-feature cross-domain matching provided by an embodiment of the present invention. The method for converting human-clothing space data by multi-feature cross-domain matching may at least include steps S100-S400:

[0053] S100, obtaining measurement data, ages, and pattern preferences of a number of individuals, and dividing the individuals into a plurality of groups according to age.

[0054] It should be noted that the type of clothing to be designed is determined based on actual conditions. In the embodiments of the present application, jeans are taken as an example. Since jeans are a classic casual clothing, they are suitable for people of all ages. Therefore, the age distribution of the audience of jeans is relatively wide, and the degree of preference for the style of jeans is also different for each person. In other embodiments, other types of clothing can be taken as examples, which do not constitute specific limitations. Therefore, the corresponding measurement data measured is static measurement data for the lower body, and the style preference is provided for individuals to choose from multiple options, such as: straight tube, tight, etc., and multiple selections can be made. Therefore, the style preference of each individual is equivalent to a word segmentation set of words including several options. Among them, the measurement data, age, and style preference can be obtained by conducting market research on brand jeans consumer groups for different waist types, different fabric elasticity, etc., or through professional measurement, questionnaire survey, Internet collection, etc., without specific limitations.

[0055] Optionally, the measurement data includes a plurality of body index data, each of which may be longitudinal data along the long axis of the lower body of the human body or transverse data on the cross section of the lower body of the human body. The longitudinal data include but are not limited to parameters such as the inseam length, the outseam length, the crotch depth, the knee circumference, the ankle circumference, etc., and the transverse data include but are not limited to parameters such as the waist circumference, the hip circumference, the thigh circumference, the front wave, the back wave, etc.

[0056] S200, determining the location significance of each body index data of each individual according to the measurement data, style preferences and the total number of individuals in all groups.

[0057] S300: Determine a target weight value for each body index data according to the site significance, and determine a target linear regression equation according to the target weight value.

[0058] S400: Determine target parameters for designing clothing according to the measurement data to be analyzed and the target linear regression equation.

[0059] The technical solution of the embodiment of the present application obtains measurement data, age and style preference of several individuals, and divides the several individuals into multiple groups according to age, wherein the measurement data includes multiple body index data; the part significance of each body index data of each individual is determined according to the measurement data, style preference and the total number of individuals in all groups, and the target weight value of each body index data is determined according to the part significance, and the target linear regression equation is determined according to the target weight value, and the corresponding target weight value is quantified based on the part significance of each body index data, so that when the target linear regression equation performs calculation and analysis on the measurement data to be analyzed, the target parameters for designing clothing can be determined more accurately, which is conducive to meeting the size requirements of clothing for people of different body shapes.

[0060] In one embodiment, in step S100, a number of individuals are divided based on age using a K-means clustering method to obtain multiple groups (or clusters), that is, the number of groups is , divided into There are groups, and the ages of individuals in each group are relatively similar, and the individuals in the corresponding group have style preferences and body characteristics corresponding to this age group.

[0061] In one implementation, step S200 includes steps S201-S202:

[0062] S201. Determine the group consistency of individuals in each group based on the measurement data and pattern preferences.

[0063] Optionally, step S201 includes S2011-S2013:

[0064] S2011. Determine the consistency of characteristics between individuals in each group based on measurement data and pattern preferences.

[0065] Optionally, based on the Jaccard coefficient and the pattern preferences of the individuals, the characteristic similarity values ​​of the pattern preferences between the individuals in each group are calculated. For example, the Jaccard coefficient is , No. In the group The individual's version preference is , No. In the group The individual's version preference is , because you can choose multiple options for your style preference, such as straight, tight, etc., , is a word segmentation set including several option words; by calculating The consistency of characteristics among individuals in each group can be determined.

[0066] Then, calculate the absolute value of the difference in each corresponding physical index data between individuals in each group. , Indicates In the group The measurement data of individuals Body index data, No. In the group The measurement data of individuals body index data; and according to the number of body index data And the absolute value of each difference , calculate the mean difference: ;

[0067] Finally, according to the exponential function and the mean difference Get the first calculation result: , according to the first calculation result Similarity value to feature The first product of , so the total formula is:

[0068]

[0069] in, is the consistency of characteristics among individuals in each group , specifically In the group Individual and The consistency of individual characteristics, The value of is (1, ], The value of is (1, -1]. Indicates The number of individuals in a population. The larger the value, the In the group Individual and The more similar the pattern preference characteristics of the individuals are, the In the group Individual and Individuals’ demands and aesthetics for jeans are similar; The smaller the value, the In the group Individual and The measurement data of individuals The more similar the body index data are. In the group Individual and The more similar the pattern preference characteristics of the two individuals are, the more similar the measurement data corresponding to the two individuals are. In the group Individual and The greater the consistency of the characteristics of the individuals.

[0070] S2012. Taking feature consistency as a distance indicator, cluster each group separately to obtain a first sub-cluster and a second sub-cluster corresponding to each group.

[0071] Optionally, the above-mentioned feature consistency is used as a distance indicator, and the K-means method is used again to cluster each group separately. All individuals in the group are further divided into two sub-groups to obtain the first sub-cluster and the second sub-cluster corresponding to each group. The number of individuals in the two sub-groups and the closeness between the individuals are then analyzed to determine the first sub-cluster. The group consistency corresponding to each group.

[0072] S2013. Determine the group consistency of individuals in each group according to the first sub-cluster, the second sub-cluster, the number of individuals in each group, and the feature consistency.

[0073] First, if the number of individuals in the first subcluster is greater than the number of individuals in the second subcluster, the first subcluster is taken as the first target subcluster and the second subcluster is taken as the second target subcluster; otherwise, the first subcluster is taken as the second target subcluster and the second subcluster is taken as the first target subcluster, that is, the subcluster with more individuals is taken as the first target subcluster and the other as the second target subcluster.

[0074] Secondly, determine the first ratio of the number of individuals in each first target sub-cluster to the number of individuals in the corresponding group ,in, Indicates The number of individuals in the first target sub-cluster divided by the population, Indicates The number of individuals in a group.

[0075] Furthermore, the target feature consistency between individuals in the first target subcluster is determined from the feature consistency , and determine the consistency mean corresponding to the target feature consistency: ,in, For the In the group Individual and The consistency of individual characteristics.

[0076] Finally, according to the normalization function and the mean consistency , determine the second calculation result , according to the second calculation result Ratio to the first The second product of determines the group consistency of individuals in each group , specifically The group consistency of individuals in a group, so the total formula is:

[0077]

[0078] in, The larger the value, the The more individuals in a group are concentrated in the first target sub-cluster, The larger the value, the The greater the feature consistency between each individual and other individuals in the first target sub-cluster of the group, the higher the degree of closeness between the individuals in the corresponding first target sub-cluster. The more individuals in the group are concentrated in the first target sub-cluster, and the higher the degree of closeness between individuals in the first target sub-cluster, the corresponding The greater the group consistency of the individuals in a group.

[0079] S202. Determine the location significance of each body index data of each individual according to the group consistency and the total number of individuals in all groups.

[0080] It should be noted that The more individuals there are in a group, the more concentrated the target group is in that group, and the greater the significance of the corresponding index of that group. However, there may be a situation where there are many individuals in the group and the individual's pattern preference and measurement data vary greatly. Therefore, it is necessary to combine the group consistency of the individuals in each group and the number of individuals in each group to determine the significance of the index of the individuals in each group.

[0081] Optionally, step S202 includes S2021-S2022:

[0082] S2021. Determine the significance of the indicators of individuals in each group based on the total number of individuals in all groups, the number of individuals in each group, the group consistency of individuals in each group, and the sum of the consistency of the group consistency of individuals in all groups.

[0083] Optionally, based on the total number of individuals in all groups ( For the The number of individuals in each group ( Different values ​​can represent the number of individuals in each group), the group consistency of individuals in each group ( The values ​​of different groups can represent the group consistency of individuals in each group) and the sum of the group consistency of individuals in all groups , determine the significance of the individual indicators in each group .

[0084] Specifically, the group consistency of the individuals in each group is determined separately The sum of the group consistencies of all the individuals in the group The second ratio ; Determine the number of individuals in each group The total number of individuals in all groups The third ratio ; According to the second ratio Ratio to the third The second product of determines the significance of the individual indicators in each group , so the total formula is:

[0085]

[0086] in, For the The significance of the individual indicators in a group, When the values ​​are different, they can represent the group consistency of individuals in each group; The larger the value, the In the group The greater the group consistency of the individuals in a group. The larger the value, the In the group The more individuals there are in a group, The larger the value, the The greater the significance of the individual indicator in the group.

[0087] S2022. Determine the location significance of each body indicator data of each individual according to the indicator significance and the body indicator data.

[0088] Optionally, the part prominence includes a first part prominence and a second part prominence.

[0089] First, the body index data of each individual constitutes a set. For example, the body index data includes longitudinal data, which includes inseam length, outseam length, crotch depth, knee circumference, and ankle circumference. That is, the longitudinal data includes indivual( =5), we get Then, the variance of each body index data is calculated for each set, and the body index data are sorted from small to large according to the size of the variance. It should be noted that the variance can reflect the clustering of the data in each set. If the clustering of the data in a set is greater, it means that the individual's change in the body index data is smaller. Since the body index data with greater changes contributes more to the explanatory power of the regression model, the smaller the clustering of the body index data in a set, the more information it can provide for the regression model, so the significance of the part corresponding to the body index data should be greater.

[0090] For example, when the body index data includes longitudinal data, the variance of the longitudinal data is calculated separately, for example, the variance of the inner seam length, outer seam length, crotch depth, knee circumference, and ankle circumference is calculated. If the corresponding variances of the inner seam length, outer seam length, crotch depth, knee circumference, and ankle circumference are variance of outer seam length > variance of inner seam length > variance of knee circumference > variance of ankle circumference > variance of crotch depth, then the arrangement results are crotch depth, ankle circumference, knee circumference, inner seam length, and outer seam length, and the corresponding arrangement positions have arrangement numbers such as 1-5; similarly, when the body index data includes transverse data, the variance is calculated using the same method and then arranged; and if the body index data includes both longitudinal and transverse data, then the longitudinal data is arranged separately after the variance is calculated, and the transverse data is arranged separately after the variance is calculated, and two arrangement results are obtained, and the two arrangement results are independent of each other.

[0091] Secondly, when the body index data is longitudinal data along the long axis of the lower body, determine the sum of the first sequence numbers of all longitudinal data. , and the sorting number of each vertical data , For the The position (serial number) of each longitudinal data in the sorting sequence, and the sorting serial number of each longitudinal data is determined respectively The sum of the first serial number The fourth ratio , calculate the fourth ratio respectively The significance of the indicators corresponding to the individuals The third product of the first site significance of each longitudinal data of each individual is obtained. , so the formula is:

[0092]

[0093] in, Indicates The individual The site significance of longitudinal data, Indicates The individual belongs to The significance of the indicators for each group; The larger the value, the The greater the significance of the position of the i-th longitudinal data of an individual, The larger the value, the The individual belongs to The greater the significance of the indicator of a group, The larger the value, the The later the position of the vertical data in the sorting sequence is, the The smaller the data clustering of the longitudinal data, Indicates the number of longitudinal data, for example, when the longitudinal data is inseam length, outseam length, crotch depth, knee circumference, ankle circumference, =5.

[0094] Similarly, when the body indicator data is the transverse data of the dimensions on the cross-section of the lower body of the human body, determine the sum of the second serial numbers of the sorting numbers of all the transverse data and the sorting number of each transverse data, determine the fifth ratio of the sorting number of each transverse data to the sum of the second serial numbers, and calculate the fourth product of the fifth ratio and the significance of the indicator corresponding to the individual to obtain the second part significance of each transverse data of each individual, which will not be repeated here.

[0095] In one embodiment, the target linear regression equation includes a target first linear regression equation and a target second linear regression equation, and step S300 includes steps S301-S303:

[0096] S301. Determine the sum of the first-part significances by summing the total number of individuals in the group and the first-part significances of each longitudinal data of each individual, and obtain the target first weight value of each longitudinal data according to the sixth ratio of the sum of the first-part significances to the total number of individuals in the group.

[0097] Optionally, based on the total number of individuals in the group And the first site significance of each longitudinal data of each individual Sum and determine the sum of the significance of the first part , and according to the sum of the first part significance The total number of individuals in the group The sixth ratio of the target first weight value of each longitudinal data is obtained. , so the calculation formula is:

[0098]

[0099] in, Indicates the first parameter that needs to be set in the preset linear regression equation. The target first weight value of the longitudinal data, The larger the value, the The larger the target first weight value of the longitudinal data is, the larger the target first weight value of the longitudinal data is.

[0100] S302, summing up the second part significance of each horizontal data of each individual according to the total number of individuals in the group, determining the sum of the second part significance, and obtaining the target second weight value of each horizontal data according to the seventh ratio of the sum of the second part significance to the total number of individuals in the group.

[0101] Similarly, based on the principle of the above formula, the target second weight value of each horizontal data that needs to be set in the preset linear regression equation can be determined.

[0102] S303: Obtain a target first linear regression equation according to the target first weight value and the preset linear regression equation, and obtain a target second linear regression equation according to the target second weight value and the preset linear regression equation.

[0103] It should be noted that the preset linear regression equation is set based on the existing method and will not be repeated here. The focus of this application is to set the original weight of the preset linear regression equation to the target first weight value of this application, thereby obtaining the target first linear regression equation, or to set the original weight of the preset linear regression equation to the target second weight value of this application, thereby obtaining the target second linear regression equation. Among them, the target first linear regression equation is used to calculate the relationship between the longitudinal data and the height value, and the target second linear regression equation is used to calculate the relationship between the transverse data and the hip circumference value.

[0104] In one embodiment, the target parameters include a height value and a hip circumference value. In step S400, after obtaining the measurement data to be analyzed, if the measurement data to be analyzed has longitudinal data to be analyzed, the longitudinal data to be analyzed is processed using the target first linear regression equation to determine the height value of the designed clothing; if the measurement data to be analyzed has transverse data to be analyzed, the transverse data to be analyzed is processed using the target second linear regression equation to determine the hip circumference value of the designed clothing.

[0105] Optionally, after determining the height and hip circumference values, the basic type of jeans can be determined by combining three-dimensional cutting and planar structure. The aesthetics and comfort of the sample pants can then be evaluated to obtain the best matching value for the jeans, which can then be applied to the research and development and production practice of branded jeans.

[0106] In the embodiment of the present application, jeans are taken as an example of clothing. In the manufacturing process of jeans, height and hip circumference play a very important role, which directly affect the size, version and final wearing effect of jeans. Therefore, by combining the index significance of individuals in each group and the importance of different body index data, the first part significance and the second part significance of each horizontal data or vertical data are set, so as to determine the corresponding target first weight value and target second weight value, determine the above-mentioned target first linear regression equation and target second linear regression equation, and finally determine the height value and hip circumference value through the above-mentioned target first linear regression equation and target second linear regression equation, which is conducive to more accurate design of jeans and to meet the size requirements of jeans for people of different body shapes.

[0107] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multi-tasking and parallel processing are also possible or may be advantageous.

[0108] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. A method for converting human-clothing spatial data by multi-feature cross-domain matching, characterized in that: The method comprises: Acquire measurement data, ages, and style preferences of a plurality of individuals, and divide the individuals into a plurality of groups according to the ages, wherein the measurement data includes a plurality of body index data; Determining the feature consistency between individuals in each group according to the measurement data and the pattern preference; the more similar the pattern preference features of two individuals in the group are, the more similar the measurement data corresponding to the two individuals are, and the greater the feature consistency between the two individuals is; Using the feature consistency as a distance indicator, clustering each of the groups respectively to obtain a first subcluster and a second subcluster corresponding to each of the groups; Determine the group consistency of the individuals in each of the groups according to the first sub-cluster, the second sub-cluster, the number of individuals in each of the groups, and the feature consistency; the more the individuals in the group are concentrated in the first target sub-cluster and the higher the degree of closeness between the individuals in the group, the greater the group consistency of the individuals in the group; Determine the site significance of each of the body index data of each of the individuals according to the group consistency and the total number of individuals in all of the groups; Determining the site significance of each of the body index data of each of the individuals according to the group consistency and the total number of individuals in all the groups includes: Determine the significance of the indicator of the individuals in each of the groups according to the total number of individuals in all of the groups, the number of individuals in each of the groups, the group consistency of the individuals in each of the groups, and the sum of the consistency of the group consistency of the individuals in all of the groups; Determining the site significance of each of the body indicator data of each of the individuals according to the indicator significance and the body indicator data; Determine a target weight value of each of the body index data according to the site significance, and determine a target linear regression equation according to the target weight value; The target parameters for designing the clothing are determined according to the measured data to be analyzed and the target linear regression equation.

2. According to the method for human-clothing spatial data conversion based on multi-feature cross-domain matching in claim 1, it is characterized by: Determining the consistency of characteristics between individuals in each group according to the measurement data and the pattern preference includes: Calculate the characteristic similarity values ​​of the pattern preferences between individuals in each group according to the Jaccard coefficient and the pattern preferences corresponding to the individuals; Calculate the absolute value of each corresponding difference in the physical indicator data between individuals in each group, and calculate the mean value of the difference according to the number of the physical indicator data and each absolute value of the difference; A first calculation result is obtained according to the exponential function and the mean difference value, and the feature consistency between the individuals in each of the groups is obtained according to the first calculation result and the first product of the feature similarity value.

3. According to the method for converting human-clothing spatial data by multi-feature cross-domain matching in claim 1, it is characterized by: Determining the group consistency of individuals in each of the groups according to the first sub-cluster, the second sub-cluster, the number of individuals in each of the groups, and the feature consistency includes: If the number of individuals in the first subcluster is greater than the number of individuals in the second subcluster, the first subcluster is used as the first target subcluster and the second subcluster is used as the second target subcluster; otherwise, the first subcluster is used as the second target subcluster and the second subcluster is used as the first target subcluster; respectively determining a first ratio of the number of individuals in each of the first target sub-clusters to the number of individuals in the corresponding group; Determining target feature consistency between individuals in the first target subcluster from the feature consistency, and determining a consistency mean corresponding to the target feature consistency; A second calculation result is determined according to the normalization function and the consistency mean, and the group consistency of each individual in the group is determined according to a second product of the second calculation result and the first ratio.

4. According to the method for converting human-clothing spatial data by multi-feature cross-domain matching in claim 1, it is characterized by: Determining the significance of the indicators of the individuals in each of the groups according to the total number of individuals in all of the groups, the number of individuals in each of the groups, the group consistency of the individuals in each of the groups, and the sum of the consistency of the group consistency of the individuals in all of the groups includes: Determining a second ratio of the group consistency of each individual of the group to the sum of the consistency of the group consistency of the individuals of all groups; Determining a third ratio of the number of individuals in each of the groups to the total number of individuals in all of the groups; The significance of the indicator of each individual in the group is determined according to a second product of the second ratio and the third ratio.

5. According to claim 1, the method for converting human-clothing spatial data by multi-feature cross-domain matching is characterized in that: Determining the site significance of each of the body indicator data of each individual according to the indicator significance and the body indicator data includes: Calculating the variance of each of the body index data respectively, and sorting the body index data from small to large according to the size of the variance; When the body index data is longitudinal data along the long axis of the lower body of the human body, determine the sum of the first sequence numbers of the sorting sequence numbers of all the longitudinal data and the sorting sequence number of each longitudinal data, respectively determine the fourth ratio of the sorting sequence number of each longitudinal data to the sum of the first sequence numbers, respectively calculate the third product of the fourth ratio and the significance of the index corresponding to the individual, and obtain the first part significance of each longitudinal data of each individual; When the body indicator data is transverse data of the dimensions on the cross-section of the lower body of the human body, determine the sum of the second serial numbers of the sorting numbers of all the transverse data and the sorting number of each of the transverse data, respectively determine the fifth ratio of the sorting number of each of the transverse data to the sum of the second serial numbers, respectively calculate the fourth product of the fifth ratio and the significance of the indicator corresponding to the individual, and obtain the second part significance of each transverse data of each of the individuals.

6. The method for converting human-clothing spatial data by multi-feature cross-domain matching according to claim 1 is characterized in that: Determining the target weight value of each of the body index data according to the part significance, and determining the target linear regression equation according to the target weight value includes: Determine the sum of the first-part significances according to the total number of individuals in the group and the first-part significances of each longitudinal data of each individual, and obtain the target first weight value of each longitudinal data according to the sixth ratio of the sum of the first-part significances to the total number of individuals in the group; Determine the sum of the second part significance according to the total number of individuals in the group and the second part significance of each horizontal data of each individual, and obtain the target second weight value of each horizontal data according to the seventh ratio of the sum of the second part significance to the total number of individuals in the group; According to the target first weight value and the preset linear regression equation, a target first linear regression equation is obtained, and according to the target second weight value and the preset linear regression equation, a target second linear regression equation is obtained.

7. The method for converting human-clothing spatial data by multi-feature cross-domain matching according to claim 6 is characterized in that: Determining the target parameters for designing clothing according to the measured data to be analyzed and the target linear regression equation includes: Determining a height value for designing clothing according to the longitudinal data to be analyzed in the measurement data to be analyzed and the target first linear regression equation; Determining a hip circumference value of the designed clothing according to the to-be-analyzed transverse data in the to-be-analyzed measurement data and the target second linear regression equation; The target parameters include height and hip circumference.

Images