Digital size standard establishment method for virtual fitting

Abstract

The invention discloses a digital size standard establishment method for virtual fitting, and the method comprises the following steps: obtaining original human body size data and a 3D human body grid model from a mobile terminal and a three-dimensional scanning device, and carrying out the unit unification, abnormal value filtering and normalization processing, thereby forming a standardized human body size data set; body type classification: based on the standardized human body size data set or the 3D human body grid model, identifying the body type category of the user through a body type classification model; based on a preset body type category and weight mapping relation, weight values are distributed for different parts, and the names, the size numerical values, the units and the weights of the parts form structured tetrad labels; and sorting according to the weight to form an ordered digital size sequence which is used for personalized recommendation and virtual fitting matching. According to the invention, the size data of the user is collected and encrypted and stored in the account, so that the user can directly load a digital size for use when purchasing the wearable article, and convenient and accurate size matching is realized.

Description

Technical Field

[0001] This invention relates to the field of virtual fitting technology, and more particularly to a method for establishing digital size standards for virtual fitting. Background Technology

[0002] Currently, the most commonly used mainstream sizing standards in China include the Chinese National Standard (GB / T1335), with the format (height) / (net chest or waist circumference) + body type code, such as 170 / 88A; the international alphabetic sizing system, with common symbols: XS (small), S (small), M (medium), L (large), XL (extra large), XXL (super large); and the European and American numerical sizing system (inches / footes + weight), 30 / 2.3 feet (110-120 catties).

[0003] Currently, mainstream sizing systems are based on the principle of range adaptation, which is not accurate enough and does not take into account the diversity of human body shapes. When people buy clothes online, they cannot try them on and can only judge the effect by looking at the models wearing the products. However, everyone's height, weight, bust, waist, and hip measurements are different, which leads to the difference between the effect on different users and the effect shown on the models after purchase. This results in problems such as returns and exchanges, which affects the customer's shopping experience. Summary of the Invention

[0004] The purpose of this invention is to provide a method for establishing digital size standards for virtual try-on. By collecting the user's own size data, transmitting it in encrypted form, and storing it in the user's account, the user can directly load and use the digital size when purchasing clothing items online or offline.

[0005] The technical solution adopted in the method for establishing digital size standards for virtual fitting disclosed in this invention is: A method for establishing digital size standards for virtual fitting includes the following steps: Human body data acquisition and standardization: Acquire raw human body size data and 3D human body mesh models from mobile terminals and 3D scanning equipment, perform unit unification, outlier filtering and normalization processing to form a standardized human body size dataset; Body type classification: Based on the standardized human body size dataset or 3D human body mesh model, the user's body type category is identified through a body type classification model; Digital size label generation: Based on the preset body type category and weight mapping relationship, weight values ​​are assigned to different parts, and the part name, size value, unit and weight are combined into a structured four-tuple label; Digital size sequence construction: Based on the priority weight configuration of body parts according to the user's body type category, the corresponding body part digital size tags are sorted from high to low weight to form an ordered structured digital size sequence for quick filtering of wearable products.

[0006] As a preferred embodiment, the body shape classification model includes a body shape rule model, a statistical / machine learning model, and an AI image classification model.

[0007] As a preferred option, the body type classification includes: X type, A type, Y type, H type, O type, and standard type.

[0008] As a preferred embodiment, the digital size tag is used to calculate the fit of the wearer by comparing the body's digital size with the size of the wearable item during virtual try-on.

[0009] As a preferred embodiment, the digital size sequence is generated based on digital size labels according to preset encoding rules, and generates body part size values ​​that include upper / lower body identifiers, body type identifiers, and weighted sorting by JSON structure or string.

[0010] As a preferred option, digital size protection is also included, which includes: end-to-end encryption of data transmission using an encryption protocol, encryption technology for data storage, and authentication for data use.

[0011] The beneficial effects of the digital size standard establishment method for virtual fitting disclosed in this invention are as follows: By acquiring and standardizing human body data, raw data from mobile terminals and 3D scanning devices are integrated. After unit unification, outlier filtering, and normalization, a high-quality standardized human body size dataset is formed. On this basis, principal component analysis or deep learning models are used to extract key body shape features, and predefined rules or trained body shape classification models are used to identify the user's body shape category, thereby achieving accurate body shape classification. Based on the preset body shape category and weight mapping relationship, weight values ​​are assigned to different parts, and structured digital size labels are generated for use when purchasing clothing and accessories. The human body digital size is compared with the clothing and accessory digital size to provide feedback on the fit of the clothing and accessories. Based on the priority weight configuration of body parts belonging to the user's body type category, digital size tags are sorted by weight to construct an ordered digital size sequence, providing a precise basis for personalized recommendations and virtual try-on matching of wearable products. In summary, this achieves efficient collection and accurate processing of human body data, accurately extracts and classifies key body features, and generates a structured digital size sequence for personalized recommendations and virtual try-on matching, while ensuring data security and significantly improving the personalized experience of wearable products and the security of data management. Attached Figure Description

[0012] Figure 1 This is a flowchart of a method for establishing digital size standards for virtual fitting according to the present invention. Detailed Implementation

[0013] The present invention will be further described and illustrated below with reference to specific embodiments and the accompanying drawings: Please refer to Figure 1 A method for establishing digital size standards for virtual fitting includes the following steps: Human body data acquisition and standardization: Acquire raw human body size data and 3D human body mesh models from mobile terminals and 3D scanning equipment, perform unit unification, outlier filtering and normalization processing to form a standardized human body size dataset; Body type classification: Based on standardized human body size datasets or 3D human body mesh models, the body type classification model identifies the user's body type category. Body shape classification models include those based on human proportion rules, statistical / machine learning models, and AI image classification models; Body type classifications include: X type, A type, Y type, H type, O type, and standard type; Based on the weight of each part in the digital size sequence, the weighted Euclidean distance is used to calculate the matching degree between the user's body shape and each pattern in the clothing pattern database, and the pattern with the highest matching degree is selected for recommendation. Digital size label generation: Based on the preset body type category and weight mapping relationship, weight values ​​are assigned to different parts, and the part name, size value, unit and weight are combined into a structured four-tuple label; Digital size sequence construction: Based on the priority weight configuration of body parts according to the user's body type category, the corresponding body part digital size tags are sorted from high to low weight to form an ordered structured digital size sequence for quick filtering of wearable products; The digital size sequence is generated based on the digital size label according to the preset encoding rules. It is generated by using JSON structure or string to include upper / lower body identification, body type identification, and body part size values ​​sorted by weight.

[0014] It also includes digital size protection, which includes: end-to-end encryption of data transmission using encryption protocols, encryption technology for data storage, and authentication required for data use.

[0015] The aforementioned collection of digital sizes enables the construction of intelligent digital wearable systems, such as digital wardrobes. By matching users' digital sizes with suitable clothing and accessories, and with the help of AI, intelligent matching recommendations are provided. This makes it easier for users to choose more suitable clothing and accessories that meet their needs, improves the utilization rate of the clothing and accessories they own, and also reduces the trouble caused by people purchasing clothing and accessories that do not fit properly.

[0016] This invention provides a method for establishing digital size standards for virtual fitting. Through human body data acquisition and standardization, raw data from mobile terminals and 3D scanning devices are integrated. After unit unification, outlier filtering, and normalization, a high-quality standardized human body size dataset is formed. Based on this, principal component analysis or deep learning models are used to extract key body shape features, and predefined rules or trained body shape classification models are used to identify the user's body shape category, thereby achieving accurate body shape classification. Based on a preset body shape category and weight mapping relationship, weight values ​​are assigned to different body parts, and structured digital size labels are generated. These labels are used to calculate the fit of clothing by comparing the human body size with the size of the clothing when purchasing it. Based on the priority weight configuration of body parts belonging to the user's body type category, digital size tags are sorted by weight to construct an ordered digital size sequence, providing a precise basis for personalized recommendations and virtual try-on matching of wearable products. In summary, this achieves efficient collection and accurate processing of human body data, accurately extracts and classifies key body features, and generates a structured digital size sequence for personalized recommendations and virtual try-on matching, while ensuring data security and significantly improving the personalized experience of wearable products and the security of data management.

[0017] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit the scope of protection of the present invention. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the essence and scope of the technical solutions of the present invention.

Claims

1. A method for establishing digital size standards for virtual fitting, characterized in that, Includes the following steps: Human body data acquisition and standardization: Acquire raw human body size data and 3D human body mesh models from mobile terminals and 3D scanning equipment, perform unit unification, outlier filtering and normalization processing to form a standardized human body size dataset; Body type classification: Based on the standardized human body size dataset or 3D human body mesh model, the user's body type category is identified through a body type classification model; Digital size label generation: Based on the preset body type category and weight mapping relationship, weight values ​​are assigned to different parts, and the part name, size value, unit and weight are combined into a structured four-tuple label; Digital size sequence construction: Based on the priority weight configuration of body parts according to the user's body type category, the corresponding body part digital size tags are sorted from high to low weight to form an ordered structured digital size sequence for quick filtering of wearable products.

2. The method for establishing digital size standards for virtual fitting as described in claim 1, characterized in that, The body shape classification model includes models based on human proportion rules, statistical / machine learning models, and AI image classification models.

3. The method for establishing digital size standards for virtual fitting as described in claim 1, characterized in that, The body type classifications include: X type, A type, Y type, H type, O type, and standard type.

4. The method for establishing digital size standards for virtual fitting as described in claim 1, characterized in that, The digital size tag is used to calculate the fit of the wearer by comparing the body's digital size with the size of the wearable item during virtual try-on.

5. The method for establishing digital size standards for virtual fitting as described in claim 1, characterized in that, The digital size sequence is generated based on digital size labels according to preset encoding rules. It is generated by using JSON structure or strings to include upper / lower body identifiers, body type identifiers, and body part size values ​​sorted by weight.

6. The method for establishing digital size standards for virtual fitting as described in claim 1, characterized in that, It also includes digital size protection, which includes: end-to-end encryption of data transmission using encryption protocols, encryption technology for data storage, and authentication required for data use.

Images