KR20230109903A - Online e-commerce system using 3D virtual fitting - Google Patents

Abstract

The present invention relates to an online e-commerce system using 3D virtual fitting, and to a virtual fitting method using a network comprising a user terminal, an operator terminal, a seller terminal, and a server computer connected through a network to perform virtual fitting, wherein a user's original image, i.e., glasses or contact lenses, is obtained through a network, and a virtual fitting image is provided through an artificial neural network so that the user can fit the desired template in a virtual environment through a terminal carried by the user.

Description

Online e-commerce system using 3D virtual fitting}

The present invention relates to an online e-commerce system using 3D virtual fitting, and more particularly, to an online e-commerce system using 3D virtual fitting, which provides a virtual user image wearing glasses and sunglasses to guide the user to optimal product information, thereby significantly improving the user's product purchasing power even online.

In general, customers who purchase fashion products such as eyeglasses or sunglasses wear the products themselves and purchase them after comprehensively considering whether or not they suit them well, and how they will look while driving a car or on the beach in summer after wearing them.

However, in order to directly wear the product, the buyer must visit the product seller directly, and thus suffers from time loss, cost loss such as transportation cost, and various inconveniences and inconveniences.

On the other hand, in recent years, various services that have been provided offline have been variously performed using the Internet.

That is, the above-described glasses, sunglasses, hats, clothes, gloves, shoes, headbands, socks, stockings, scarves, accessories, fashion products, and fashion products such as jewelry can be easily purchased through an online homepage, that is, electronic commerce.

However, the services provided by these conventional online e-commerce sites are limited to just introducing products, so they cannot provide various information about products to consumers, and there is a problem in that they cannot give consumers the same feeling as if they were actually wearing the products themselves.

An object of the present invention to solve the above problems is to provide an online e-commerce system using 3D virtual fitting that can provide various and realistic information about products when purchasing various fashion products such as eyeglasses and sunglasses, as well as give the feeling of a standard model or user himself wearing the product and looking in a mirror, giving a feeling when a third person looks at it, or providing various virtual fitting information and virtual experience information, such as the scenery that can be experienced while wearing the product.

In order to achieve the above object, the virtual fitting method using a network of the present invention includes at least one mobile terminal, a PDA, a smartphone, a user terminal including a camera, a camcorder, a laptop computer, a personal computer, an operator terminal, and a server computer connected through a network to perform virtual fitting. In the virtual fitting method using a network, the server computer comprises an artificial neural network, a control unit for controlling a program, and a database for storing various information, the artificial neural network includes a generation network and a classification network, and the control unit includes a user authentication program for authenticating a user, A product image information generation program for generating product image information from a product, a virtual fitting program for generating virtual fitting image information by receiving user image information and combining it with product image information, and a virtual fitting image information transmission program for transmitting the generated virtual fitting image information to the user terminal, wherein the database includes a user authentication information database in which user authentication information is stored, a product image information database in which the product image information is stored, a user image information database in which the user image information is stored, and a virtual fitting image information database in which the virtual fitting image information is stored, via (a) a network Providing product image information to a user terminal; (B) receiving user image information through a user terminal; (c) generating virtual fitting image information obtained by combining selected product image information with the user image information; and (d) transmitting the virtual fitting image information to a user terminal through a network.

The control unit further includes a model wearing image information generating model wearing image information obtained by combining standard model image information for each gender, age, body type, and face type inputted in advance with product image information, and the database further includes a standard model image information database for storing standard model image information and a model wearing image information database for storing model wearing image information. It is possible to further include; transmitting to the terminal.

As described above, according to the online e-commerce system using the 3D virtual fitting of the present invention, it is possible to provide various and realistic optimal product information about a product, thereby increasing consumers' purchasing power for the product, and providing virtual fitting information and virtual experience information as if going to a store and purchasing a product directly, so that more accurate and diverse information can be provided to the user.

1 is a conceptual diagram showing an online e-commerce system using 3D virtual fitting according to a preferred embodiment of the present invention.
2 is a conceptual diagram illustrating an exchange relationship between an operator operating an online e-commerce system using a 3D virtual fitting of the present invention, a seller selling a product, and a user.
3 is a diagram for explaining the detailed configuration of an artificial neural network according to an embodiment of the present invention.
4 is a diagram for explaining the detailed configuration of a generation network according to an embodiment of the present invention.
5 is a diagram for explaining input data and output data of a generation network according to an embodiment of the present invention.
6 is a diagram for explaining the detailed configuration of a global segmentation network according to an embodiment of the present invention.
7 is a diagram for explaining the detailed configuration of a regional division network according to an embodiment of the present invention.
8 is a block diagram showing a 3D virtual fitting method of the present invention.
9 is a flowchart illustrating an example of the 3D virtual fitting of FIG. 8 in more detail.

Prior to the detailed description of the present invention, the terms or words used in the present specification and claims described below should not be construed as being limited to the usual or dictionary meanings, and the inventors should interpret their inventions in the best way. Based on the principle that the concept of terms can be appropriately defined, they should be interpreted as meanings and concepts consistent with the technical spirit of the present invention. Therefore, the embodiments described in this specification and the configurations shown in the drawings are only the most preferred embodiments of the present invention, and do not represent all of the technical spirit of the present invention, so that they can be substituted at the time of the present application. It should be understood that there may be equivalents and variations.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. At this time, it should be noted that the same components in the accompanying drawings are indicated by the same reference numerals as much as possible. In addition, detailed descriptions of well-known functions and configurations that may obscure the gist of the present invention will be omitted. For the same reason, some components in the accompanying drawings are exaggerated, omitted, or schematically illustrated, and the size of each component does not entirely reflect the actual size.

First, as shown in FIG. 1, a virtual fitting system using a network according to a preferred embodiment of the present invention largely includes a user terminal 500, an operator terminal 600, a seller terminal 700, and a server computer 10 connected through the network 20.

Here, the server computer 10 is programmed to provide product image information to the user terminal 500 through the network 20, receive user image information from the user terminal 500, generate virtual fitting image information by combining product image information with the user image information, and transmit the result to the user terminal 500, and is a computer provided in a company that provides a high-end virtual fitting service through the network 20 in exchange for product sales fees and commissions from the user.

In addition, the user terminal 500 is a computer of a prospective purchaser or a prospective purchaser company who intends to purchase a product, and is a computer of an individual or company that receives a virtual fitting service and a virtual experience service.

Here, the user terminal 500 is not necessarily a computer, but can receive various image information and select various commands, such as various mobile terminals, PDAs, smart phones, cameras, camcorders, notebooks, personal computers, etc. It can be included.

In addition, the user terminal 500, the operator terminal 600, the vendor terminal 700, and the server computer 10 of FIG. 1 are connected to each other through the network 20, and the terminals connected by the network 20 have an Internet browser (Netscape, Internet Explorer, etc.) capable of displaying web contents in the form of HTML, XML, etc., or other internal or external, or local / long-distance wired/wireless network access protocol devices.

Meanwhile, the server computer 10 includes an artificial neural network 100, a control unit 200 for controlling programs, and a database 300 for storing various types of information.

In particular, the control unit 200 is connected to a 2D camera, 3D camera, various 3D graphic devices, 2D/3D scanners, etc. to output or generate 2D still image information, moving picture information, 3D graphic information, etc. for the product.

That is, as shown in FIG. 1, the control unit 200 includes a main program for operating all programs, a user authentication program for authenticating a user, a product image information generation program for generating product image information from a product, a virtual fitting program for generating virtual fitting image information by receiving user image information and combining it with product image information, a virtual fitting image information transmission program for transmitting generated virtual fitting image information to the user terminal, combining standard model image information for each gender, age, body type, and face type previously input into product image information A model wearing program that generates model wearing image information, a size adjustment program that receives user size information and adjusts user image information to match product size information, a 3D image conversion program that converts 2D image information of a user into 3D image information and combines it with 3D image information of a product, a virtual tilting program that converts virtual fitting image information according to an angle to generate virtual tilting image information, a product sales program that receives purchase information from the user terminal 500 and generates product sales information, and the user can see the selected product with the naked eye It may include a virtual haptic program and other programs that generate haptic virtual haptic information.

Here, the main program operates all programs, can be expressed in the form of a home page online, and can control all programs described later by receiving a command signal from the user terminal 500 or operator terminal 600.

In addition, the user authentication program authenticates a user for the purpose of purchasing a predetermined user information, fee, or product, and is a program that allows an ID or password to be input through the user terminal 500 or automatically connected to the server computer 10 through the network 20.

In addition, the product image information generation program generates product image information from a product, and is a program capable of generating 2D still image information, video information, 3D graphic information, etc. for the output product by being connected to a 2D camera or 3D camera connected to the operator computer 20, various 3D graphic devices, 2D/3D scanners, etc.

In addition, the virtual fitting program generates virtual fitting image information by receiving user image information and combining it with product image information. Although it is possible to generate virtual fitting image information simply by overlapping 2-dimensional product image information with 2-dimensional user image information, preferably, it is a program capable of generating 3-dimensional user image information based on 2-dimensional user image information and virtually combining it with pre-input 3-dimensional product image information to generate virtual fitting image information.

Also, the virtual fitting image information transmission program is a program that transmits generated virtual fitting image information to the user terminal 500 .

In addition, the model wearing program generates model wearing image information by combining standard model image information for each gender, age, body type, and face type inputted in advance with product image information.

In addition, the size adjustment program receives user size information and adjusts the user image information to match the product size information. The user directly inputs the size information or infers the size in consideration of the scale included in the user's image, and applies it to adjust the user image information to match the product size information.

In addition, the 3D image conversion program converts the user's 2D image information into 3D image information and combines it with the 3D image information of the product. It is a program that combines the 3D image information of the product in consideration of contrast, shade, shooting angle, shooting distance, etc. from a 2D still image or video.

In addition, the virtual tilting program converts virtual fitting image information according to an angle to generate virtual tilting image information, and by changing the position of an observer viewing the three-dimensional image information, it is a program that generates virtual tilting image information as if looking at a user wearing a product from various angles.

In addition, the product sales program is a program that generates product sales information by receiving purchase information of the user terminal 500 and transmits it to the operator terminal 600 and/or the seller terminal 700 so that the product can be sold.

In addition, the virtual experience program generates virtual experience information that the user can experience with the naked eye when wearing a selected product, as if the user wears a product such as glasses and / or sunglasses, and then looks at the view or surroundings. It is a program that can feel the background.

In addition, the controller 200 may further include various other programs such as various product sales fees or cost payment programs.

On the other hand, the database 300 includes a user authentication information database for storing user authentication information, a product image information database for storing product image information, a user image information database for storing user image information, a virtual fitting image information database for storing the virtual fitting image information, a standard model image information database for storing standard model image information, a model wearing image information database for storing model wearing image information, a user size information database for storing user size information, a product size information database for storing product size information, and a converted three-dimensional image information of users and products. It may include a 3D image information database, a virtual tilting image information database for storing the virtual tilting image information, a product sales information database for storing the product sales information, and a virtual experience information database for storing virtual experience information.

Here, the user authentication information is information necessary for user authentication by the user authentication program, and may include information related to authentication such as a user password, ID, IP address, authentication date, and authentication effective date.

In addition, the product image information is provided by an operator or a seller by the product image information generation program, and may include information on all products that can be worn by a user, such as glasses, sunglasses, hats, clothes, gloves, shoes, headbands, socks, stockings, scarves, accessories, fashion products, and jewelry.

At this time, it is preferable that the product size or 3D image information is described or stored in detail in the product image information.

In addition, the user size information is provided by the user through the size adjustment program, and may include numerical information directly input by the user, enlargement/reduction information directly operated by the user, automatic measurement information that is relatively automatically measured by photographing together with a scale member such as a standard ruler, coin, cigarette pack, cup, ballpoint pen, etc.

Therefore, when the user takes a picture with the scale member, the size of the user's image information can be automatically adjusted according to the product image information without the need to directly input the body size.

In addition, the 3D image information, generated by the 3D image conversion program, includes single still image conversion 3D image information that converts user's 2D image information into 3D image information by recognizing the direction of light and difference in contrast, multiple still image conversion 3D image information that converts 2 or more users' 2D image information taken at different angles into 3D image information by combining 2D image information of two or more users taken at different angles, and 3D image information that converts moving image information of users taken from various angles into 3D image information. It may include video conversion, 3D image information, and the like.

Here, in order to obtain such 3D image information, a 2D camera installed in the user terminal 500 as well as a 3D camera installed in the user terminal 10 or a specific place can be used.

In addition, when the product image information includes glasses and/or sunglasses information, the virtual fitting image information is obtained by combining the product image information and user image information by the virtual fitting program, and the color of the glasses or sunglasses, the shape of the lens, the number of lenses, the weight or thickness of the lens, the material of the lens, the coating state of the lens, the color of the glasses frame, the shape of the glasses frame, the weight or thickness of the glasses frame, the material of the glasses frame, the structure of the glasses nose, the color of the nose, the shape of the nose, the weight or thickness of the nose, the glasses nose It may include information such as the material of the eyeglasses and the structure of the eyeglass nose.

Accordingly, the user can freely replace the lenses, frames, and noses of glasses and sunglasses in a virtual state.

In addition, the virtual sensory experience information allows the user to experience virtual sensory reality by the virtual sensory experience program, and may include surrounding views such as street/outside/mountain/sea/beach viewed from a car when wearing the selected glasses and/or sunglasses, ambient views according to morning/afternoon/time, ambient views according to seasons such as summer/fall/winter, and comparative views before/after wearing the glasses and/or sunglasses.

Accordingly, the user can experience as if wearing virtually selected glasses and sunglasses, and thus obtain information on optimal glasses and sunglasses suitable for his or her purpose and intention.

In addition, the virtual tilting information is to give the user a feeling of observing product image information, user image information, and virtual fitting image information from various angles up and down, front and back, left and right, by the virtual tilting program.

Accordingly, the user can view the product image, the user image, and the user image wearing the virtually fitted product by, for example, rotating left and right or tilting up and down.

Therefore, as shown in FIG. 2, the operator operating the server computer 10 can provide users with high-quality products and virtual fitting services, and can generate revenue by receiving product sales fees and commissions, and the user with the user terminal 10 can thereby reduce the time and cost required to directly wear the product.

In addition, the operator can generate revenue by receiving a high-quality product or commission in exchange for providing the seller with an opportunity to generate revenue, and the seller can provide the user with a quality product and receive a product sales fee to obtain an opportunity to generate revenue. This can benefit all users, operators, and sellers.

Here, various types of business are possible, such as the operator may also serve as a seller, or the operator sells products in place of the seller.

3 is a diagram for explaining the detailed configuration of an artificial neural network according to an embodiment of the present invention. Referring to FIG. 3 , the artificial neural network 100 includes a generation network 110 and a classification network 120 . In addition, the segmentation network 120 includes a global segmentation network 121 and a regional segmentation network 123 . Each of the generation network 110, the global segmentation network 121, and the regional segmentation network 123 constitutes one artificial neural network. Each of the generation network 110, the global segmentation network 121, and the regional segmentation network 123 will be described.

First, the generation network 110 will be described. 4 is a diagram for explaining the detailed configuration of a generation network according to an embodiment of the present invention. 5 is a diagram for explaining input data and output data of a generation network according to an embodiment of the present invention.

Referring to FIGS. 4 and 5 , the generator network 110 includes a plurality of layers including a plurality of operations to which weights are applied. Here, the plurality of layers including a plurality of operations includes at least one convolution layer (CL) that performs a convolution operation, a pooling layer (PL) that performs a down sampling operation, and an unpooling layer (UL) layer that performs an upsampling operation and one or more deconvolution layers (DL) that performs a deconvolution operation. Each of convolution, downsampling, upsampling, and deconvolution operations uses a filter composed of a predetermined matrix, and values of elements of this matrix become weights.

The generating network 110 receives the template 1, which is an image of glasses or contact lenses, and the original non-wearing image 20, which is an image in which the user does not wear glasses or contact lenses. In particular, the original non-wearing image 20 includes pixel values of each pixel and location information of each pixel. Here, the location information has a predetermined reference point (eg, the focus of a camera) and is a 3-dimensional coordinate value (x, y, z) representing a 3-dimensional distance from the reference point (0, 0, 0) to an object photographed with a corresponding pixel. For example, any one pixel P1 of the original non-wearing image 20 is basically located at (x, y) according to the pixel coordinate system, has pixel values (eg, RGB values), and has three-dimensional coordinate values (x, y, z).

When the template 1 and the original non-wearing image 20 are input, the generation network 110 performs a plurality of calculations to which weights of a plurality of layers are applied to the original non-wearing image 20 to generate a virtual wearing image 30, which is an image in which the user wears glasses or contact lenses. In other words, the generation network 110 generates a virtual wearing image 30, which is an image in which the user wears glasses or contact lenses, by performing a plurality of operations in which weights of a plurality of layers are applied to pixel values and 3D coordinate values of all pixels of the original non-wearing image 20. Accordingly, each pixel of the generated virtual wearing image 30 includes location information as well as a pixel value. The location information of the virtual wearing image 30 is a three-dimensional coordinate value (x, y z) having the same reference point (eg, camera focus) as the original non-wearing image 20 . Accordingly, it is possible to derive the template specification 2 representing the actual size of glasses or content lenses from the virtual wearing image 30 generated by the generating network 110 . In other words, in an image captured by a camera (digital camera), parameters of the camera, such as focal length, focal length, resolution, and image sensor specifications, are stored in metadata. In addition, a distance from a reference point (eg, a focus of a camera) to a corresponding object may be known through location information. Accordingly, the horizontal and vertical actual sizes of the image can be obtained through the focal length, the principal point distance, and the location information, and the actual size of each pixel can be obtained through the resolution. When the virtual wearing image 30 is generated from the original non-wearing image 20, metadata is inherited as it is. Therefore, the actual size and position of each pixel of the virtual wearing image 30 can be obtained using the metadata and location information of the virtual wearing image 30 . Accordingly, if the actual size of each pixel is arranged according to location information (three-dimensional coordinate values), a template standard (2) representing the actual size of eyeglasses or content lenses of the actual size can be derived. The original unworn image 20 according to an embodiment of the present invention is photographed using a camera (digital camera) and a 3D sensor. At this time, the camera photographs the image and generates a pixel value for each pixel, and the 3D sensor measures the distance from a predetermined reference point to the object constituting each pixel of the image to generate location information from the measured distance. The virtual wearing image 30 also has location information according to the same criterion. For example, any one pixel P2 of the virtual wearing image 30 represents a spectacle frame and has a three-dimensional coordinate value. As described above, the template specification (2) can be derived using camera parameters and location information.

Next, the division network 120 including the global division network 121 and the regional division network 123 will be described. 6 is a diagram for explaining the detailed configuration of a global segmentation network according to an embodiment of the present invention. 7 is a diagram for explaining the detailed configuration of a regional division network according to an embodiment of the present invention.

As shown in FIGS. 6 and 7 , both the global segmentation network 121 and the local segmentation network 123 include a plurality of layers that perform a plurality of operations to which weights are applied. Here, the plurality of layers that perform a plurality of operations include a convolution layer (CL) that performs a convolution operation and a fully connected layer (FL) that performs a soft-max operation. The convolution operation uses a filter composed of a predetermined matrix, and values of elements of this matrix become weights. In addition, the softmax operation is also performed by applying weights.

The global segmentation network 121 receives the wearing images 30 and 40, which are images of the user wearing glasses, and performs a plurality of calculations to which weights are applied to the wearing images 30 and 40, and outputs whether the input wearing images 30 and 40 are real or fake. Here, the wearing images 30 and 40 include a virtual wearing image 30 and an original wearing image 40 . The virtual wearing image 30 is an image generated by the generating network 110, and the original wearing image 40 is an original image of a user wearing glasses.

The regional segmentation network 123 receives local images 35 and 45, performs a plurality of calculations to which weights are applied to the local images 35 and 45, and outputs whether the input regional images 35 and 45 are real or fake. Here, the region images 35 and 45 include a virtual wearing region image 35 and an original wearing region image 45 . The virtual wearing region image 35 is extracted from the virtual wearing image 30 and means a rectangular image of a predetermined size including eyes, noses, ears, glasses or contact lenses identified in the virtual wearing image 30 . The original wearing region image 45 is extracted from the original wearing image 40 and means a rectangular image of a predetermined size including eyes, noses, ears, glasses or contact lenses identified in the original wearing image 40 .

As described above, both the global segmentation network 121 and the regional segmentation network 123 perform convolutional operations. The convolution operation is performed using a filter (or kernel), which is a matrix of a predetermined size, and features of an image are extracted from the convolution operation.

On the other hand, as shown in FIG. 8, if the virtual fitting method using a network according to a preferred embodiment of the present invention is sequentially shown, the virtual fitting method using a network according to a preferred embodiment of the present invention comprises: constructing a server computer that performs the virtual fitting method using the network of the present invention, (a) providing product image information to the user terminal 10 through the network 20; (b) receiving user image information through the user terminal 500; Generating virtual fitting image information by combining selected product image information with image information; (d) transmitting the virtual fitting image information to the user terminal 500 through the network 20; (e) transmitting the virtual experience information to the user terminal 500 through the network 20; It can be made including the step of doing.

Here, the step (a) may further include receiving product image information and standard model selection information from the user terminal 500 and transmitting model wearing image information obtained by combining the selected product image information with the standard model image information to the user terminal 500 through the network 20.

Further, the step (b) may further include receiving user size information from the user terminal 500 and adjusting user image information to match product size information.

In addition, the step (c) may further include converting the user's 2D image information into 3D image information and combining it with the 3D image information of the product.

Further, the step (d) may further include receiving angle selection information from the user terminal and transmitting the virtual tilting image information to the user terminal through a network.

If an example of these steps (a) (b) (c) (d) (e) is shown in more detail using a flowchart, as shown in FIG. 9, when a user accesses the server computer (S1) and inputs an ID and password to receive user authentication (S2), various product image information is provided to the user terminal 500 so that the user can check the product image (S3).

Subsequently, when the user selects at least one or more product images from among the checked product image information (S4), the selected product image information is prepared (S5).

At this time, the prepared product image information may include both a two-dimensional product image and a three-dimensional product image as well as the detailed size of the product.

Next, when the user selects, for example, a standard model for virtual fitting, such as a domestic or foreign model or celebrity that fits or matches the user's gender, body type, age, face shape, etc. (S6), the standard model image desired by the user can be prepared (S7).

Alternatively, when the user inputs a user image as a model without selecting a standard model (S8), the user image can be prepared using a camera installed in the user terminal 500 (S9).

At this time, the user is asked to input the user's size (S10), and the user's image size can be adjusted to match the product image (S11).

Then, if the user wants 3D conversion to check the product and user's image from various angles (S12), the product image and user image can be converted into 3D images, respectively, and prepared (S13).

If a two-dimensional virtual fitting is desired, for example, when only the front image of glasses is overlapped on the user's front face in two dimensions, or when a three-dimensional image of various angles is prepared, the user selects virtual fitting (S14). The prepared two-dimensional or three-dimensional images are combined to generate a virtual fitting image (S15), and provided to the user terminal 500 so that the user can check the virtual fitting image (S16).

Subsequently, if the user wants to virtually tilt the virtual fitting image so that the user can enjoy the virtual fitting image from various angles by adjusting the "3D adjustment button" (S17), the 3D image is angle-converted to create a virtual tilting image (S18), and allows the user to check it (S19).

Subsequently, when the user wants a virtual experience of the scenery, background, or surrounding objects while wearing actual glasses or sunglasses (S20), virtual experience information is generated in consideration of the color or transparency of the sunglasses lens selected by the user (S21), and the user's virtual experience information can be checked (S22).

Therefore, the virtual fitting method using the network of the present invention provides various and realistic information about the product by providing a screen that a standard model or user directly wears and looks in a mirror or a third person enjoys in purchasing various fashion products such as eyeglasses or sunglasses, as well as various virtual fitting information and virtual experience information such as the scenery that can be experienced while wearing the product.

Therefore, from the operator's point of view, it is possible to promote sales by increasing the purchasing power of the product by providing customized high-quality information to the user, and from the user's point of view, it is possible to virtually experience both the actual appearance and feeling of wearing the product, thereby minimizing the purchase of the product incorrectly or being disappointed with the product, and receiving a high-quality service that allows the user to enjoy the optimal shopping environment even online.

Meanwhile, the present invention can also be implemented as computer readable codes in a computer readable recording medium.

Computer-readable recording media include all types of recording devices in which data that can be read by a computer system is stored.

Examples of computer-readable recording media include ROM, RAM, CD-ROM, magnetic tape, floppy disk, and optical data storage devices, and also include those implemented in the form of carrier waves (for example, transmission through the Internet).

In addition, the computer-readable recording medium is distributed to computer systems connected through a network, so that computer-readable codes can be stored and executed in a distributed manner.

The present invention is not limited to the above-described embodiments, and modifications by those skilled in the art are possible within a range that does not impair the spirit of the present invention.

For example, in the embodiment of the present invention, the product seems to be limited to glasses or sunglasses, but the products are not necessarily limited to glasses or sunglasses, but glasses, sunglasses, hats, clothes, gloves, shoes, headbands, socks, stockings, scarves, accessories, fashion products, jewelry, etc. All products that can be worn by the user can be applied.

Therefore, the scope claimed in the present invention will not be determined within the scope of the detailed description, but will be limited by the claims described later and their technical spirit.

10: server computer 20: network
100: artificial neural network 110: generative network
120: segmentation network 121: global segmentation network
123: regional division network 200: control unit
300: database 500: user terminal
600: operator terminal 700: seller terminal

Claims (2)

A virtual fitting method using a network comprising a user terminal, an operator terminal, a seller terminal, and a server computer connected through a network to perform virtual fitting,
The server computer is composed of an artificial neural network, a controller for controlling a program, and a database for storing various information,
When the artificial neural network receives a template and an original wearing image from the user terminal, a wearing generation network that generates a virtual non-wearing image, which is an image of the user not wearing glasses or contact lenses, from the received original wearing image, and a wearing-generation network that generates a virtual wearing image, which is an image of the user wearing the template, from the generated virtual non-wearing image;
The control unit includes a user authentication program for authenticating a user, a product image information generation program for generating product image information from a product, a virtual fitting program for generating virtual fitting image information by receiving user image information and combining it with product image information, and a virtual fitting image information transmission program for transmitting generated virtual fitting image information to the user terminal,
The database includes a user authentication information database in which user authentication information is stored, a product image information database in which the product image information is stored, a user image information database in which the user image information is stored, and a virtual fitting image information database in which the virtual fitting image information is stored. According to claim 1,
The control unit further includes a 3D image conversion program that converts the user's 2D image information into 3D image information and combines it with the 3D image information of the product,
The database further comprises a three-dimensional image information database for storing three-dimensional image information of converted users and products.