KR20230083440A - Non-face-to-face customized item matching system and method - Google Patents

Abstract

Non-face-to-face customized item matching system and method according to the embodiment matches fashion items such as registered glasses to the user based on user information such as face shape and face color extracted through user taste and 3D image analysis, and consumers You can check the result of matching the item to your face and purchase it. In addition, in the embodiment, user information and product information are analyzed based on big data to recommend products more suitable to the user's taste, and sales and purchases of products such as glasses can be performed non-face-to-face.

Description

Non-face-to-face customized item matching system and method {NON-FACE-TO-FACE CUSTOMIZED ITEM MATCHING SYSTEM AND METHOD}

The present disclosure relates to an item matching system and method, and specifically, to a system and method for optimally recommending fashion items such as glasses, hats, earrings, and necklaces to consumers in a non-face-to-face manner.

Unless otherwise indicated herein, material described in this section is not prior art to the claims in this application, and inclusion in this section is not an admission that it is prior art.

In about a year after the outbreak of COVID-19, our daily lives and society as a whole have completely changed. To enter an indoor space or building, you have to pass through a thermal sensor, and body temperature measurement has become a must. Each of them has to put an extra layer of veil on their face and go out into the streets. It's just a thin mask, but the social and psychological distance is much farther than that. As 'social distancing' became the best way to prevent the spread of infection, social activities and daily life such as classes, work, and food orders have changed to non-face-to-face methods.

After a long winter vacation, elementary, middle and high schools started a new semester with a first-of-its-kind 'online lecture', and online digital took over the submission of assignments and simple tests to prove attendance. Numerous companies have attempted to work from home, and employees have flocked to video conferencing. The customer center, which had evaluated that it would not be possible to work from home due to concerns about personal information leakage, also introduced a cloud system and started working from home. As it became 　virtually 　difficult to watch on-site performances, the style of watching performances has changed a lot from before. The non-face-to-face society is rapidly becoming deeply embedded in our lives.

Our society shaken by COVID-19 has also brought about major changes in consumption patterns. The daily life of shopping at traditional markets or hypermarkets is giving way to non-face-to-face consumption due to the spread of COVID-19. As social distancing is strengthened and going out decreases, the paradigm of the food service industry is shifting to delivery. In particular, as social distancing is strengthened, eating in cafes is prohibited in the metropolitan area, and it is impossible to eat at general restaurants after 9:00 pm, the demand for delivery is further increasing. Delivery agency company B said, “The number of delivery orders from January 1 this year to the 13 of this month was 123 million, a 132% increase compared to the same period last year.” Online malls are also enjoying a boom. According to the National Statistical Office, online shopping transaction amount in October was 14.2 trillion won, up 20.0% from a year ago. As the focus of consumption shifts to online, offline businesses are also strengthening their online business. The duty-free industry, which suffered from a sharp drop in airport passengers, has been selling duty-free products in stock through online malls since last June, and large fashion brands have also contributed to expanding their sales channels online, such as building online-only logistics centers. are taking

Non-face-to-face shopping and online malls are rapidly expanding, but fashion items such as glasses, clothes, shoes, necklaces, earrings, etc. must be worn and purchased by consumers themselves or purchased products that match the physical characteristics of consumers, so fashion items such as glasses There are many consumers who return or are dissatisfied with the purchase after ordering non-face-to-face.

1. Korean Patent Registration No. 10-1566707 (2015.11.02) 2. Korean Patent Registration No. 10-2255404 (2021.05.17)

Non-face-to-face customized item matching system and method according to the embodiment matches fashion items such as registered glasses to the user based on user information such as face shape and face color extracted through user taste and 3D image analysis, and consumers You can check the result of matching the item to your face and purchase it.

In addition, in the embodiment, user information and product information are analyzed based on big data to recommend products more suitable to the user's taste, and sales and purchases of products such as glasses can be performed non-face-to-face.

The non-face-to-face customized item matching system according to the embodiment collects body information including face and body image information, body size, and visual acuity, and preferred item details including brand, color, price, and style of the preferred item. smart terminal; a store terminal registering detailed information and item image information of products provided in the store and items available for sale; a matching server that converts image information into 3D and recommends items according to user body information and preferred item detail information; includes

A non-face-to-face customized item matching method according to another embodiment includes (A) image information of a face and body photographed on a smart terminal, body information including body size and visual acuity, and brand, color, price, and style of preferred items. Collecting preferred item details and transmitting them to a matching server; (B) registering detailed information of products provided in the store and sellable items in the store server and transmitting the information to the matching server; (C) The matching server collects the consumer's body information, preferred item details, and sellable item details for each store, converts the image included in the body information into 3D, and converts the user's body information and preferred item details analyzing; and (D) recommending an item to the consumer according to the analysis result in the matching server; includes

The non-face-to-face customized item matching system and method as described above provides fashion items such as glasses, shoes, necklaces, and earrings to users based on the 3D image analysis results including the user's face shape and facial color information and the user taste analysis results. Since the recommendation is made, it is possible to improve the purchase satisfaction of the fashion item, which needs to consider the user's physical characteristic information. In addition, after matching items such as eyeglasses with their own 3D face image in advance, consumers can check the matching result and decide to purchase the product, thereby reducing the return rate of products such as eyeglasses.

The effects of the present invention are not limited to the above effects, and should be understood to include all effects that can be inferred from the detailed description of the present invention or the configuration of the invention described in the claims.

1 is a diagram showing the configuration of a non-face-to-face customized item matching system according to an embodiment
2 is a diagram showing a data processing block of a matching server according to an embodiment
3 is a diagram showing a data processing configuration of a generation module according to an embodiment
4 is a view showing an embodiment of combining an item and a consumer's body image;
5 is a signal flow diagram of a non-face-to-face customized item matching system according to an embodiment

Advantages and features of the present invention, and methods for achieving them, will become clear with reference to the embodiments described below in detail in conjunction with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but may be implemented in various different forms, and only the present embodiments make the disclosure of the present invention complete, and common knowledge in the art to which the present invention belongs It is provided to fully inform the holder of the scope of the invention, and the present invention is only defined by the scope of the claims. Like reference numerals designate like elements throughout the specification.

In describing the embodiments of the present invention, if it is determined that a detailed description of a known function or configuration may unnecessarily obscure the subject matter of the present invention, the detailed description will be omitted. In addition, terms to be described below are terms defined in consideration of functions in the embodiments of the present invention, which may vary according to the intention or custom of a user or operator. Therefore, the definition should be made based on the contents throughout this specification.

1 is a diagram showing the configuration of a non-face-to-face customized item matching system according to an embodiment.

Referring to FIG. 1 , a non-face-to-face customized item matching system according to an embodiment may include a smart terminal 100 of a consumer, a matching server 200, and a store terminal 300. The consumer's smart terminal 100 receives the consumer's body information and preferred item detailed information. In an embodiment, the consumer's body information may include information such as height, weight, face shape, skin color, eyesight, etc. input from the consumer, and image information of the face and body. The preferred item details may include information such as brand, color, price, style, and material of the item preferred by the consumer.

The store terminal 300 registers detailed information on products provided in the store and items available for sale. In the embodiment, the detailed information of the item may include meta data of the sale item, such as item code, brand, cost, selling price, discount rate, whether or not there is an event.

The matching server 200 collects the consumer's body information, preferred item details, and item detail information from the consumer's smart terminal 100 and the store terminal 300, analyzes them, and recommends an optimal item to the consumer. For example, the matching server 200 may convert an image transmitted from a consumer's smart terminal into a 3D image, and recommend an item matching the user's body information and preferred item detail information.

2 is a diagram showing data processing blocks of a matching server according to an embodiment.

Referring to FIG. 2 , the matching server according to the embodiment may include an information collection module 210 , an analysis module 220 , a generation module 230 and a recommendation module 240 . The term 'module' used in this specification should be interpreted as including software, hardware, or a combination thereof, depending on the context in which the term is used. For example, the software may be machine language, firmware, embedded code, and application software. As another example, the hardware may be a circuit, processor, computer, integrated circuit, integrated circuit core, sensor, micro-electro-mechanical system (MEMS), passive device, or combination thereof.

The information collection module 210 collects body information and preferred item details input by the consumer from the consumer's smart terminal, and collects the consumer's face and body image information captured by the smart terminal. In addition, the information collection module 210 collects detailed information on items that can be sold in each store from the store terminal.

The analysis module 220 analyzes the collected consumer's body information, image information, preferred item details, and item details. In an embodiment, the analysis module 220 may analyze body information, image information, preferred item details, and item details collected by using a data mining technique. Data mining is a process of finding useful information that is not easily revealed in a vast amount of data, a technique for finding hidden relationships between data, or relationships that are not revealed or are too complex to be extracted through existing statistical methods. This is based on all available data, including daily transaction data, customer data, product data, or other external data other than customer response data of various marketing held by the company, to discover hidden knowledge, unexpected trends or new rules. It is actively used to discover information and use it as information for actual business decision-making. The data mining technique used in the embodiment analyzes the degree of association between items and customers through steps of sampling, exploration, modification and adjustment, modeling, and assessment.

Data mining is based on large amounts of data ranging from tens of megabytes to tens of gigabytes. However, looking at a huge amount of data is a task that requires a lot of patience and effort, even in terms of time. Therefore, sampling is to extract a small amount of data that resembles a population from a large amount of data (population). Sampling is an essential procedure that can reduce the time and cost of data mining work, help efficient work progress, and lighten the weight of data mining work through this. Random sampling, stratified sampling, etc. are commonly used for sampling. Data extracted through sampling undergoes exploration. The process of data exploration is a step of examining the relationship between numerous variables that we have, starting with confirming and quantifying the facts we already know. At this stage, what was actually known as one population can be divided into two or more, dozens or more variables can be reduced to a small number of variables that give actual important information, and new variables can be created in addition to the existing variables. .

The transformation and modification process is a step to efficiently use the information possessed by the data to improve the performance of the model in the modeling step based on the information obtained in the exploration step. There are methods such as variable transformation, quantification, and grouping that create new variables or modify existing variables in consideration of tasks in which new variables can be created using existing variables.

The modeling process can be said to be the core of data mining, and based on the results from the previous step, it provides methods for finding a predictive model through an appropriate technique according to the purpose of analysis. Models used in this stage include Neural Network, CHAID (Chi-Automatic Interaction Detection), CART (Classification And Regression Trees), and generalized linear models. The evaluation process is to select the best model in terms of reliability and usefulness by comparing the effects of two or more models derived in the modeling process. The selected model can be reflected in the actual population to re-evaluate its effect, and if a new problem is raised due to this, the previous work can be repeated again.

In an embodiment, the analysis module 220 extracts data samples by sampling information collected from smart terminals and store terminals as a population. In addition, by referring to the information provided from the external database, variables necessary for calculating the relationship between consumers and items are derived. At this time, the variables target consumer information, image information, preferred item details, and item details collected from the matching server, and among the derived variables, only the most effective variables for each consumption are selected again to establish the relationship between consumers and items. yield Association refers to a relationship between items hidden in data, and analyzing such a relationship is referred to as association analysis or link analysis. In most cases, a relatively simple rule is extracted, such as 'If A happens, B happens', for example, 'If you buy a hammer, you also buy a nail' at a hardware store. In addition, a relationship specifically related to the flow of time is also referred to as a sequence rule. A representative technique used for this purpose is market basket analysis, which is derived from identifying purchasing behavior by analyzing items in a customer's shopping cart. Among the methods of analyzing correlation, connection analysis visually expresses the relationship between phenomena in which mutual correlation exists, thereby making it possible to access the hidden structure and linkage of data in various and easy-to-understand ways, and exploring the properties of data. Based on this, it is a method to easily discover interesting patterns or new patterns.

The generating module 230 creates a combined image. In the embodiment, the combined image is an image obtained by combining items such as glasses, hats, earrings, and necklaces with a user's image converted into a 3D image. In the embodiment, the generated combined image is transmitted to the smart terminal and the store server, and the consumer can check the combined image in which the item is combined with the consumer's image and make an item purchase decision.

The recommendation module 240 recommends items including eyeglasses and hats in consideration of user's body analysis information and body information including facial shape and color extracted through 3D image analysis of the user. In addition, the recommendation module 240 may recommend items that the consumer may prefer in consideration of the combined image and the relationship data between the consumer and the item transmitted from the analysis module.

In addition, the recommendation module 240 analyzes user body information, user body analysis information, user preferred item details, item details, and user purchase history information based on big data, and recommends items to the user based on the analysis results. , so that the sale of recommended items can be carried out non-face-to-face.

3 is a diagram showing a data processing configuration of a generation module according to an embodiment.

Referring to FIG. 3 , the generation module according to the embodiment may include a scanning unit 231, a feature extraction unit 232, and a coupling unit 233. The scanning unit 231 scans a user's face or body image and an item image. In an embodiment, the scanning unit 231 may perform 3D scanning to shape the consumer's face, body, and items obtained from a smart terminal or store terminal in 3D. The feature extractor 232 extracts features of the scanned consumer's face and body images and item images. For example, a registered item image is scanned to extract features for each item, and a user feature is extracted by scanning an image of a user's face or body. Specifically, the feature extraction may include edge detection, color detection, size detection, feature ratio detection, eye size, eyebrow length, eyebrow-to-eye length, and pupil distance detection processes. In addition, item feature extraction may include a process of detecting the size, curvature, shape, and color of the item.

The coupling unit 233 combines an item selected by the consumer or an image of an item recommended by the analysis server to have a relational relationship of a certain level or higher with the corresponding consumer's body image. Also, in an embodiment, the coupler 233 couples the item image to the consumer's body image in consideration of the wearing position of the item and the consumer's body characteristics. For example, when the image of glasses is combined with the image of the consumer's face, the center line of the glasses and the center line of the consumer's face are combined so that they match, and when the image of the earring item is combined, the image of the earring item is combined with the wearing position of the earlobe of the consumer adjust to be In addition, when the hat item is combined with the consumer's face image, the distance between the center line of the hat and the center line of the consumer's face image may be adjusted to be less than a predetermined value.

4 is a diagram illustrating an embodiment of combining an item and a consumer's body image.

Referring to FIG. 4 , in the embodiment, an image of a user's face is combined with an image of a fashion item such as glasses or a hat so that the user can confirm that the item is worn on the user's face and decide to purchase the corresponding item. As shown in FIG. 4 , in the embodiment, since a combined image is generated by replacing the glasses 10 or the hats 30 and 40 coupled to the consumer's body image 20 and provided to the consumer, the consumer himself/herself You can purchase the item that suits you after checking the combination image.

Hereinafter, a non-face-to-face customized item matching method will be described in turn. Since the operation (function) of the non-face-to-face customized item matching method according to the embodiment is essentially the same as the function of the non-face-to-face customized item matching system, overlapping descriptions with FIGS. 1 to 4 will be omitted.

5 is a signal flow diagram of a non-face-to-face customized item matching system according to an embodiment.

Referring to FIG. 5 , in the embodiment, the consumer's body information and preferred item detailed information are registered from the smart terminal in step S10. The consumer's body information may include the consumer's face and body images photographed by a camera. In step S20, information registered in the smart terminal is transmitted to the matching server, and in step S30, the matching server collects consumer body information and preferred item details. In step S40, detailed information of items that can be sold is registered in the store terminal, and transmitted to the matching server in step S50. In step S60, detailed information on sellable items is collected from the matching server, and the information collected in step S70 is analyzed. Step S70 may include a process of analyzing a relationship between consumer body information, detailed information about a preferred item, and detailed information on a sellable item, or a process of analyzing a consumer's face and body image and an item image. In step S80, the matching server recommends an item to the consumer according to the information analysis result. In step S80, a combined image combining the consumer image and the item may be generated, and the item may be recommended to the consumer based on the generated combined image and the analysis result of the relationship between the consumer and the item. In step S90, the matching server transmits item recommendation information to the smart terminal, and in step S100, the smart terminal receives the recommendation information and purchases the recommended item. In step S110, the store terminal may sell the corresponding item upon receiving an item purchase message from the smart terminal.

The non-face-to-face customized item matching system and method as described above provides fashion items such as glasses, shoes, necklaces, and earrings to users based on the 3D image analysis results including the user's face shape and facial color information and the user taste analysis results. Since the recommendation is made, it is possible to improve the purchase satisfaction of the fashion item, which needs to consider the user's physical characteristic information. In addition, after matching items such as eyeglasses with their own 3D face image in advance, consumers can check the matching result and decide to purchase the product, thereby reducing the return rate of products such as eyeglasses.

The disclosed content is only an example, and can be variously modified and implemented by those skilled in the art without departing from the subject matter of the claim claimed in the claims, so the protection scope of the disclosed content is limited to the specific It is not limited to the examples.

Claims (7)

In the non-face-to-face customized item matching system,
A smart terminal that collects body information including face and body image information, body size and eyesight, and preferred item details including brand, color, price, and style of the preferred item;
a store terminal registering detailed information and item image information of products provided in the store and items available for sale;
a matching server that converts the image information into 3D and recommends an item according to user body information and preferred item detailed information; Non-face-to-face customized item matching system that includes.
According to claim 1, wherein the matching server; Is
A non-face-to-face customized item matching system characterized by generating a combined image by combining the image of a user converted into three dimensions with an item image including glasses, and transmitting the generated combined image to a smart terminal and a store terminal.
According to claim 1, wherein the matching server; Is
A non-face-to-face customized item matching system that recommends items including glasses and hats in consideration of body information including eyesight and body information including user body analysis information including face shape and face color extracted through 3D image analysis of the user .
According to claim 1, wherein the matching server; Is
Analyzing user body information, user body analysis information, user preference item details, item details, and user purchase history information based on big data, recommending items to users based on the analysis results, and comparing sales of recommended items A non-face-to-face customized item matching system characterized in that it proceeds face-to-face.
According to claim 1, wherein the matching server; Is
A non-face-to-face customized item matching system characterized by extracting item features by scanning a registered item image and extracting user features by scanning an image of a user's face or body.
In the non-face-to-face customized item matching method,
(A) The smart terminal collects image information of the face and body, body information including body size and visual acuity, and preferred item details including the brand, color, price, and style of the preferred item, and transmits them to the matching server step;
(B) registering detailed information of products provided in the store and sellable items in the store server and transmitting the information to the matching server;
(C) The matching server collects the consumer's body information, preferred item details, and sellable item details for each store, converts the image included in the body information into 3D, and converts the user's body information and preferred item details analyzing;
(D) recommending an item to a consumer according to the analysis result in a matching server; Non-face-to-face customized item matching method comprising a.
The method of claim 6, wherein the step of (C); Is
generating a combined image by combining an item image including glasses with a user's image converted into 3D; and
Transmitting the generated combined image to a smart terminal and a store terminal; Non-face-to-face customized item matching method comprising a.

Images