Clothing image diversification and user participation type clothing image transaction platform service providing method, device and system based on video processing

Abstract

One embodiment of the present invention is a method of providing product image diversification and user participated garment image transaction platform services that allow wholesale and retail stores operating a garment shopping center to distinguish actual photos of models uploaded to their shopping center garment product introduction page from other shopping centers, and ensuring ownership by purchasing the actual wearing photo of the model apparel product and on this basis filling in a photo having a composition, posture or corrected state different from other shopping centers, with respect to the device and system.

Description

TECHNICAL FIELD

[0001] The following embodiments show how to provide an image processing-based clothing image diversification and user-involved clothing image transaction platform service in which a wholesaler and retailer operating a clothing mall can purchase a photo of a model actually wearing a clothing product and upload it to a clothing product introduction page of their mall to distinguish from other shopping malls, secure ownership, and compose their own product introduction page by filling in the photo with a composition, pose, and correction state that is distinguished from other shopping malls. This is about the device and system. BACKGROUND

[0002] A retailer who purchases clothing from a wholesaler such as Dongdaemun Market and Sinchang Market sells the clothing using their shopping center or smart store's website.

[0003] In this case, a product introduction / description page explaining a wearing photo or product details usually uses an image provided by a wholesaler, and if all retailers use such an image, the product introduction pages between retailers will be similar, and the differentiation of each retailer will disappear.

[0004] Even if a retailer hires a model, a photographer, a web designer, etc., or uses a freelancer to distinguish themselves, the result is not satisfactory in many cases compared to the cost involved.

[0005] So far, there has not been any service that sells, trades, or distributes these "clothing photos" to retailers, and clothing retailers take photos themselves or use photos distributed in batches by wholesalers.

[0006] Therefore, there is a need for a platform that allows retailers to create their own differentiated product introduction pages without incurring additional manpower or high costs.

[0007] PRIOR ART DOCUMENT

[0008] PATENT DOCUMENT

[0009] (Patent Document 1) KR 10-2300660

[0010] (Patent Document 2) KR 10-2455412

[0011] (Patent Document 3) KR 10-2424760 SUMMARY

[0012] PROBLEMS TO BE SOLVED BY THE INVENTION

[0013] An embodiment of the present invention aims to provide a method, device and system for providing an image processing-based clothing image diversification and user participation clothing image transaction platform service, which generates wearing photos of clothes according to images taken and uploaded by models, and retailers (clothing manufacturers) can select and purchase them and build their own product introduction pages, to overcome the limitations of the above-mentioned traditional clothing shopping centers.

[0014] Means for solving the problem

[0015] According to one embodiment, the device comprises a processor, a memory, a communication module and a non-transitory storage medium, and is executed by the processor to perform the image processing-based clothing image diversification and user participation clothing image transaction platform service providing method by executing the program stored in the non-transitory storage medium, to receive the first actual image of the first piece of clothing from the model terminal; based on the above actual image, create the first real-time image; upload the first actual image to the image transaction page; according to the above image transaction page, receive the request for purchasing the image from the terminal of the clothing supplier; transmit the second actual image corresponding to the image purchase request to the terminal of the clothing manufacturer; and delete the second actual image from the image transaction page; it provides an image processing-based clothing image diversification and user participation clothing image transaction platform service providing method.

[0016] Furthermore, after removing the second real picture from the picture transaction page, the step of creating the first real picture is re-executed; the step of creating the first real picture is: calculating the first number of the third real picture corresponding to the first piece of clothes uploaded to the picture transaction page; calculating the third number by subtracting the first number from the specified second number; generating the number of the third image for the fourth time; specifying the fourth real image as the first real image; the step of creating the fourth real image is: the first generation step of converting the third real image into a vector to produce the first vector; the second generation step generates a 5-1 real image that captures frames from the real image and adds it to the 0.5 times 3 in the first image list; the third generation step converts the 5-1 real image contained in the first image list into a vector to create a second vector; the fourth generation step compares each second vector with all first vectors and produces a first similarity; the fifth generation step extracts a third vector, i.e. a second vector whose first similarity is greater than a first reference value; the sixth generation step deletes the first image from the above list; the seventh generation step generates a 5-3 real image that captures frames from the real image and generates a second image list by generating the above-mentioned third number; the eighth generation step converts the 5-4 real image contained in the above-mentioned second image list into a vector to create a fourth vector; the ninth generation step compares the fourth vector with all first and second vectors and produces a first similarity; the tenth generation step extracts a fifth vector, i.e. a fourth vector whose first similarity is less than a first reference value; the eleventh generation step adds a 5-5 real image, i.e. a 5-4 real image corresponding to the fifth vector, to the first image list and specifies the fifth vector as a second vector; the twelfth generation step initializes the second image list and repeats the seventh to eleventh generation steps until the number of elements in the first image list is greater than the number of elements in the third image list; it can be used to include.

[0017] The first generation step, the third generation step and the eighth generation step are: extracting a first color code specific to a pixel from an input image, and generating a color list with the first color code as an element; generating a 5-dimensional vector for all elements contained in the color list in turn (red as the first color, green as the first color, blue as the first color, center x coordinate and CBD y coordinate); and returning a sum vector of the 5-dimensional vector; the step of creating a five-dimensional vector is: generating a first partial image containing only the first color code from the input image; specifying the average value of the x coordinate of each pixel of the image first part code as the center x coordinate; and specifying the average value of the y coordinate of each pixel encoded in the first color code pixel of the image first part as the city center y coordinate; it can be used to include.

[0018] In addition, the image transaction page is: reading the Identifier of the clothing factory terminal; first practice reading the image; creating a viewer image with a watermark on the first actual image; and transmitting the viewer image to the clothing supplier terminal; the step of creating the viewer image is: creating a first layer and using the identifier at the top of the image for first actual use; calculating the first length, which is the average font thickness of the identifier indicated in the first layer; creating a second layer, copying the pixels from the image below the opaque pixels indicated in the first layer for first actual use; moving the 2-1 layer to the left by the first length; creating a second layer, copying the pixels from the image below the opaque pixels indicated in the 2-1 layer for first actual use; moving the 2-2 layer to the right by 2 times the first layer length; creating a second layer to a third layer, copying the pixels from the image below the opaque pixels indicated in the first layer for first actual use; 2-3 layer mosaicking; deleting the first layer; merging the above 2-1 layer, 2-2 layer and 2-3 layer into a 3-1 layer; converting the border area located outside the range of the first length from the opaque pixels indicated in the 3-1 layer to the first color; creating a 3-2 layer, copying the pixels from the image below the first color indicated in the 3-1 layer for first actual use; color inverting the 3-2 layer; returning a viewer image with the 3-2 layer on top, the 3-1 layer in the center, and the first actual use image at the bottom; it can be used to include.

[0019] In addition, the actual image is an image taken from a 360-degree direction around the model position, and the second and seventh steps are: reading the first capture angle and the first capture time of the previous capture frame from the actual image; according to the frequency of each angle of the first shooting angle, 2-1 shooting angle corresponds to the first multiple number of appearances, and 2-2 shooting angle corresponds to the second multiple number of appearances; according to the frequency of each angle of the first capture angle, 2-3 shooting angles corresponding to the first level of the lowest appearance and 2-4 shooting angles corresponding to the second level of the lowest appearance are extracted; according to the frequency of each hour of the first capture time, 2-1 capture time corresponding to the first level of the highest number of appearances and 2-2 capture time corresponding to the second level of the highest number of appearances are extracted; according to the frequency of each angle of the first capture time, 2-3 capture time corresponding to the first level of the lowest appearance and 2-4 capture time corresponding to the second level of the lowest appearance are extracted; the median of the first frame corresponding to 2-1 capture angle and 2-2 capture angle, and the median of 2-1 capture time and 2-2 capture time; the median of 2-1 capture angle and 2-2 capture angle, and the median of 2-3 capture time and 2-4 capture time corresponding to the second frame; the median of 2-3 capture angle and 2-4 capture angle, and the median of 2-1 capture time and 2-2 capture time corresponding to the third frame; the median of 2-3 capture angle and 2-4 capture angle, and the median of 2-3 capture time and 2-4 capture time corresponding to the fourth frame; the third capture angle of the fifth frame is added to the first capture angle of the fifth frame; the third capture time of the fifth frame is added to the first capture time of the fifth frame; and the fifth frame is returned, that is, any one of the first to fourth frames; it can be used to include.

[0020] According to one embodiment, the apparatus can be controlled by a computer program stored on a medium in combination with hardware to perform any of the above methods.

[0021] Inventive effect

[0022] According to one embodiment, by generating a large number of images based on the 360-degree lens shot by the model, retailers can encourage them to use different images on the product introduction page.

[0023] In addition, retailers can choose and purchase photos that match their shopping center concept or image, and own these photos to create their own product introduction pages.

[0024] In addition, by registering only the pictures that have been sold or the pictures that have a low similarity with other pictures already registered on the platform, it is possible to prevent the sale of pictures with too high a similarity.

[0025] In addition, by vectorizing the images and judging the similarity based on the vectors, it is possible to determine the similarity between images without using high-load algorithms such as artificial intelligence models.

[0026] In addition, on the platform (web page) where the image is sold, a watermark is placed on top of the image to prevent the retailer (clothing company) from using the photo without purchasing it, for example, by capturing.

[0027] In addition, from the photos that have been registered or sold on the platform, you can add photos that have different compositions but the same pose, photos that have the same composition but different poses, and photos that have different compositions and poses. BRIEF DESCRIPTION OF DRAWINGS

[0028] Figure 1 is a schematic diagram of the clothing image diversification and user participation clothing image transaction platform service providing system based on image processing in an embodiment of the present application.

[0029] Figure 2 is a flowchart of the clothing image diversification and user participation clothing image transaction platform service providing based on image processing provided according to an embodiment of the present application.

[0030] Figure 3 is a schematic diagram of the actual image shooting process describing the clothing image diversification and user participation clothing image transaction platform service providing method based on image processing according to an embodiment of the present application.

[0031] Figure 4 is an attachment of a partial image of the clothing image diversification and user participation clothing image transaction platform service providing method based on image processing according to an embodiment of the present application.

[0032] Figure 5 is a watermarking example of the clothing image diversification and user participation clothing image transaction platform service providing method based on image processing according to an embodiment of the present application.

[0033] REFERENCE NUMERALS

[0034] S100: Receiving a video for actual use

[0035] S200: Step of creating an image for the first real life

[0036] S300: Step of uploading the first real picture to the picture transaction page

[0037] S400: Step of receiving an image purchase request from a clothing retailer terminal

[0038] S500: Step of transmitting the second real image to the clothing manufacturer terminal

[0039] S600: Step of deleting the second real image from the picture transaction page

[0040] 1: Device

[0041] 2: model terminal

[0042] 3: clothing contractor terminal

[0043] 4: file server

[0044] M: model number

[0045] C: camera

[0046] I: input image

[0047] P1: first partial image

[0048] L31: 3-1 layer

[0049] answer: border area

[0050] L32: 3-2 layer DETAILED DESCRIPTION

[0051] Hereinafter, embodiments are described in detail with reference to the accompanying drawings. However, the embodiments can be variously changed and, therefore, the scope of the patent application is not limited to the embodiments. Any change, modification or replacement of the embodiments should be construed as included in the scope of the patent application.

[0052] The detailed structure or function description of the embodiments is initiated only for the purpose of illustration and can be changed and implemented in various forms. Therefore, the embodiments are not limited to the specific form of disclosure, and the scope of the specification includes changes, unification or replacement included in the descriptive concept.

[0053] Terms such as first or second can be used to describe various components, but the interpretation of these terms should be used only to distinguish one component from another component. For example, the first component can be named the second component, and similarly, the second component can be named the first component.

[0054] When a component is referred to as "connected" to another component, it is understood that it can be directly connected to or connected to another component, but there can be another component in between.

[0055] The terms used in the embodiments are only for the purpose of illustration and should not be interpreted as limiting. The singular expression includes the plural expression unless the context clearly indicates otherwise. In this specification, the term "include" or "have" should be understood as specifying the presence of features, numbers, steps, actions, components, parts or combinations thereof described herein, and does not exclude the presence or addition of one or more other features or numbers, steps, actions, components, parts or combinations thereof.

[0056] Unless otherwise defined, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art to which the embodiments belong. Terms defined in commonly used dictionaries should be interpreted as having a meaning that is consistent with their meaning in the context of the relevant description and should not be interpreted in an idealized or overly formal sense unless expressly so defined in the present application.

[0057] In addition, in describing the drawings, the same reference numerals are assigned to the same components regardless of drawing codes, and the same repetitive description is omitted. In describing the embodiments, if it is judged that a detailed description of the relevant notification technology can unnecessarily obscure the gist of the embodiments, detailed explanation is omitted.

[0058] The embodiments can be implemented in various types of products, such as personal computers, server computers, and cloud servers.

[0059] According to one embodiment, the device 1 includes a processor, a memory, a communication module, and a non-transitory storage medium, and is executed by the processor by executing a program stored in the non-transitory storage medium to provide a service of a clothing image diversification and user participation clothing image transaction platform based on image processing, receiving a first actual image of a piece of clothing from a model M terminal 2 S100; based on the above type of image, the first actual image S200; uploading the first actual image to an image transaction page S300; based on the above image transaction page, receiving a picture purchase request from a clothing manufacturer terminal 3 S400; Step S500 transmits a second actual image corresponding to the image purchase request to a clothing distributor terminal 3; and deleting the second actual image from the image transaction page S600; it provides a method for providing a service of a clothing image diversification and user participation clothing image transaction platform based on image processing.

[0060] The model M terminal 2 and the clothing manufacturer terminal 3 can be a specified smart device (smartphone, tablet) or PC.

[0061] The terms "real life image" and "real life image" used in the specification of the present application can not only refer to videos / photos taken when the model (person) wears the clothing, but also videos / photos of bottom clipping, transparent hanger clipping, human model clipping, etc. of clothing / fashion items. For example, when the clothing / fashion item is sitting on a transparent hanger, a real video can be recorded based on a large number of cameras installed around it.

[0062] The first actual image of the piece of clothing is received from the model M terminal 2 S100; based on Figure 3 As shown by the 360-degree camera C, the model receives an actual image file that simultaneously captures 360 degrees of the environment around the model M.

[0063] At this time, the actual wearing image corresponds to the first piece of clothing selected and worn by the model M, and can be stored in the file server 4. Thus, a plurality of real wearing images photographed by a plurality of models M for the same piece of clothing can be stored, and thus, the first piece of clothing image photographed by a different model M for the same first piece of clothing can be traded.

[0064] Based on the above type of image, the first actual image S200; only a few frames of the actual image (i.e., the video file) are extracted and created and saved as the first actual image.

[0065] Here, the frame refers to a "clip" that constitutes a video. For example, if 12 cameras C photograph 60 FPS for 5 minutes, as shown in Figure 3 There are 12x 60x 5 = 3600 frames in theory, which can be converted / generated / extracted into each first actual image.

[0066] The first actual use image is uploaded to the image trading page S300; the first piece of clothing, the first actual wearing corresponding image is uploaded to the image trading page (platform / web page).

[0067] At this time, the original file of the image is used for the first actual use, and is stored separately in the file server 4, and based on this, only the viewer image with a watermark can be configured to be viewable on the terminal 3 of the clothing manufacturer.

[0068] The clothing retailer views the first wearing image corresponding to the first wearer they want to sell on their device, sends a purchase request for the first wearer image they want to purchase, and pays the fee.

[0069] Based on the above image trading page, the step of receiving the image purchase request from the clothing manufacturer terminal 3 S400; here, we receive the image purchase request (purchased the first actual use image) received from the clothing manufacturer terminal 3.

[0070] Step S500 transmits the second actual use image corresponding to the image purchase request to the clothing distributor terminal 3; in the second actual use image corresponding to the image purchase request (image purchased by the clothing supplier), the second actual image (image purchased by the clothing manufacturer) is transmitted so that it can be downloaded from the terminal 3 of the clothing retailer.

[0071] Delete the second actual image from the image trading page S600; in order to prevent the second actual image that has already been purchased by someone from being resold on the image trading page (platform), the second actual image that has been sold will be deleted from the image trading page (not marked as the first actual image).

[0072] In addition, after the second actual image is deleted from the image transaction page S600, the step of creating the first actual use image S200 will be re-executed; the step of generating the first actual picture S200 is: calculating the first number of the third actual picture corresponding to the first clothes, which has been uploaded to the image transaction page; calculating the third number by subtracting the first number from the specified second number; generating the fourth actual image equal to the number of the third image above; specifying the fourth actual image as the first real image; the step of creating the fourth actual use image is: the first generation step of converting the third actual use image into a vector to produce the first vector; the second generation step generates a 5-1 real image that captures frames from the actual image and adds it to the first image list with a number of 0.5 times 3; the third generation step converts the 5-1 real image contained in the first image list into a vector to create a second vector; the fourth generation step compares each second vector with all first vectors and produces a first similarity; the fifth generation step extracts the third vector, i.e. the second vector whose first similarity is greater than the first reference value; the 5-2 actual use image, i.e. the 5-1 actual use image corresponding to the third vector, is deleted from the above list in the sixth generation step; the seventh generation step generates a 5-3 actual image that captures frames from the actual image and generates a second image list by generating the above-mentioned third number; the eighth generation step converts the 5-4 real image contained in the above-mentioned second image list into a vector to create a fourth vector; the ninth generation step compares the fourth vector with all first and second vectors and produces a first similarity; the tenth generation step extracts the fifth vector, i.e. the fourth vector whose first similarity is less than the first reference value; the eleventh generation step adds the 5-5 real image, i.e. the 5-4 real use image corresponding to the fifth vector, to the first image list and specifies the fifth vector as the second vector; the twelfth generation step initializes the second image list and repeats the seventh to eleventh generation steps until the number of elements in the first image list is greater than the number of elements in the third image list; it can be used to include.

[0073] According to one embodiment of the present application, the number of images of the first actual wearing of the first clothes that can be identified by the clothes manufacturer can be constantly maintained.

[0074] For example, if the number of the second item is specified as 50, 50 first real wearing images will be maintained for each first clothes. Therefore, if the clothes retailer purchases 10 out of the 50 first real wearing pictures (second real wearing pictures), 10 out of the 50 pictures will be deleted, and therefore another 10 [10 first real wearing images will be created and added.

[0075] After removing the second utility image from the mirror transaction page S600, re-executing the first utility image S200; create some images that have been sold and deleted for the first actual use and add them to the image transaction page.

[0076] Calculate the first number of "third actual use images corresponding to the first garment" uploaded to the image transaction page before; calculate the first number, the number of goods purchased by the garment manufacturer, and deduct it from the deleted number (the remaining third-party image number).

[0077] Calculate the third number by subtracting the first number from the specified second number; calculate the second number, the value specified by the server administrator, that is, the third number, the difference between the first number minus the first number.

[0078] Generate the fourth actual image equal to the number of the third image above; generate as many images as the number of the third image (newly added images).

[0079] For example, if the clothing retailer purchases 15 pictures when the second number is specified as 50, the remaining 35 pictures can be added to create 15 fourth pictures for the first use, a total of 50 pictures.

[0080] Specify the fourth actual image as the first real image; specify the newly created 4th real image as the 1st real image and process it the same way.

[0081] The first generation step is to convert the third actual use image into a vector to produce the first vector; in the following steps, the third real image is converted into a five-dimensional vector value according to the specific steps described below.

[0082] The second generation step generates 5-1 actual images captured from the actual images and adds them to the first image list, the number of which is 0.5 times the third number; capture frames equal to 0.5 times the third number from the actual images.

[0083] Then, add it to the first image list and continue to execute the subsequent steps.

[0084] The third generation step converts the 5-1 real-world images contained in the first image list above into vectors to create a second vector; convert all 5-1 real-world images contained in the first image list into a second vector.

[0085] The fourth generation step compares each second vector with all first vectors and produces a first similarity; based on the vectors, compare all 5-1 real-world images contained in the first image list with the third real image to calculate the first similarity.

[0086] Here, the first similarity between two vectors can be calculated by calculating the distance between the five-dimensional coordinate values of the vectors (coordinates of the end points of the vectors) and dividing them by a certain reference distance (percentage).

[0087] For example, if the distance between one vector A and another vector B is 120 and the reference distance is 150, 120 / 150 = 80% can be calculated as having the first similarity.

[0088] The fifth generation step extracts a third vector, the first similarity of the second vector being greater than a first reference value; filters out the third vector having the first similarity greater than the first reference value.

[0089] For example, if the first reference value is designated as 95%, only vectors having the first similarity greater than or equal to can be extracted.

[0090] The sixth generation step deletes the 5-2 actual use image, i.e., the 5-1 actual use image corresponding to the third vector, from the first image list; deletes the image from the image list having a high similarity to the first image.

[0091] In other words, an image having a high similarity to an image already uploaded to the image transaction page is not added as an image of the first real thing.

[0092] The seventh generation step generates the 5-3 actual image by capturing a frame from the actual image and generates the second image list by generating the above-mentioned third number; in a subsequent second iteration, the actual image will be recaptured to create more images for 5-3 actual use.

[0093] The eighth generation step converts the 5-4 real images contained in the above-mentioned second image list into vectors to create the fourth vector; converts the newly captured second image list into vectors during the iteration process.

[0094] The ninth generation step compares the fourth vector with all the first and second vectors and produces the first similarity; compares the newly captured and converted fourth vector with the first vector corresponding to the previously uploaded image and the second vector corresponding to the image already added to the first image list.

[0095] The tenth generation step extracts the fifth vector, i.e., the fourth vector having the first similarity less than the first reference value; filters out only the fifth vector having no similarity to other images registered or generated previously.

[0096] In the 11th generation step, the 5-5 actual image, i.e., the 5-4 actual image used for the 5th vector, is added to the list of first images, and the 5th vector is designated as the second vector; the images used for the 5-5 are added to the image list, which are not similar to other images previously registered or created.

[0097] The 12th generation step initializes the second image list, and the 7th to 11th generation steps are repeated until the number of elements in the first image list is greater than the number of elements in the third image list; the above steps are repeated until the first image list is completed (until the number of captured images is insufficient).

[0098] The first, third, and eighth generation steps are: extracting a first color code specific to a pixel from an input image I, and generating a color list with the first color code as an element; sequentially generating a 5-dimensional vector (red as the first color, green as the first color, blue as the first color, center x coordinate, and CBD y coordinate) for all elements contained in the color list; and returning a sum vector of the 5-dimensional vectors; the step of creating a five-dimensional vector is: creating a first partial image P1 containing only the first color code from the input image I; designating the average value of the x coordinate of each pixel encoded in the image P1 first part as the center x coordinate; and designating the average value of the y coordinate of each pixel encoded in the first partial image P1 as the CBD y coordinate; it can be used to include.

[0099] The first, third, and eighth generation steps are: they can be run in the same way based on the defined function / program / method.

[0100] For example, in the first generation step, the 3rd actual image is processed as the input image I, and the returned sum vector becomes the first vector, in the third generation step, the 5-1 actual image is processed as the input image I, and the returned sum vector becomes the second vector, in the eighth generation step, the 5-4 actual image is processed as the input image I, and the returned sum vector becomes the fourth vector.

[0101] Extracting a first color code specific to a pixel from an input image I, and generating a color list with the first color code as an element; extracting the color code (hexadecimal code) of each pixel from the input image I, which is in the form of displaying multiple primary color codes on the pixel.

[0102] Then, on this basis, it generates a color list without repeated elements.

[0103] generate a 5-dimensional vector for all elements contained in the color list in order (first color for red, first color for green, first color for blue, center x coordinate, and CBD y coordinate); in this case, separate the three colors (RGB) of the first color code, expressed in base 10, and assign them to three of the five dimensions of the vector, and assign the city center calculated from the first part of the image P1 composed of the first color code to the remaining two dimensions to calculate / generate a five-dimensional vector.

[0104] For example, the 5-dimensional vector generated from #123456 is generated as (18 (16), 52 (34), 86 (56), center x coordinate, city center y coordinate).

[0105] Return the sum vector of the five-dimensional vector; return the sum vector of the 5-dimensional vector as the output value of the input value.

[0106] For example, if 15 five-dimensional vectors I with 15 one-dimensional color codes are generated from the input image I, the sum vector of all 5 five-dimensional vectors is returned.

[0107] Generate the first part image P1, which contains only the first color code from the input image I; the input image I is divided into the first part image P1, which contains only the first color code based on color filtering.

[0108] For example, if there are 25 primary color codes that make up the input image I, 25 first part images P1 will be generated.

[0109] Figure 4 (b) in FIG. 4 shows the first part image P1 extracted from the input image I of (a) in FIG. 4.

[0110] Designate the average value of the x coordinate of each pixel of the image P1 first part encoding as the center x coordinate; in the first part P1 of the image, add the x coordinates of each opaque pixel and divide the average value of the x coordinate by the number of opaque pixels to calculate, which is designated as the x coordinate of the city center.

[0111] Designate the average value of the y coordinate of each pixel marked with the first color code of the first part image P1 as the city center y coordinate; after adding all the y coordinates of each opaque pixel contained in the first part image P1, calculate the average value of the y coordinate by dividing it by the number of opaque pixels, and designate it as the y coordinate of the city center.

[0112] Further, the image transaction page is: reading the identifier (IDentifier) 3 of the garment manufacturer terminal; first practicing reading the image; creating a viewer image with a watermark on the first actual image; and transmitting the viewer image to the garment supplier terminal 3; the step of creating the viewer image is: creating the first layer, and using the identifier at the top of the image for the first actual use; calculating the first length, which is the average font thickness of the identifier indicated in the first layer; creating the second layer, copying the pixels from the image located at the bottom of the opaque pixels indicated in the first layer for the first actual use; moving the 2-1 layer to the left by the first length; creating the second layer, copying the pixels from the image located at the bottom of the opaque pixels indicated in the 2-1 layer for the first actual use; moving the 2-2 layer to the right by 2 times the first layer length; creating the second layer to the third layer, copying the pixels from the image located below the opaque pixels indicated in the first layer for the first actual use; 2-3 layer mosaicking; deleting the first layer; merging the 2-1 layer, 2-2 layer and 2-3 layer to create a 3-1 layer L31; converting the border area A located outside the first length range from the opaque pixels indicated in the 3-1 layer L31 to the first color; creating a 3-2 layer L32, which copies the pixels from the image located at the bottom of the first color indicated in the 3-1 layer L31 for the first actual use; inverting the chroma of the 3-2 layer L32; and returning a viewer image in which the 3-2 layer L32 is located at the top, the 3-1 layer L31 is located in the center, and the image for the first actual use is located at the bottom; it can be used to include.

[0113] Reading the identifier (IDentifier) of the garment manufacturer terminal 3; here we read the predetermined identifier from the garment manufacturer terminal 3.

[0114] For example, the account ID of the garment distributor and the MAC address of the garment manufacturer terminal 3 can be received as the above-mentioned identifier.

[0115] First practicing reading the image; in the middle, the first actual wearing image corresponding to the first garment to be viewed by the garment supplier is read from the file server 4.

[0116] Creating a viewer image with a watermark on the first actual image; in the file, the file read from the file server 4 has a watermark to create a predetermined viewer image.

[0117] The step of transmitting the viewer image to the garment manufacturer terminal 3; in the first actual image, the image for actual use will not be sent directly to the terminal 3 of the garment manufacturer, but only the viewer image will be sent to the terminal 3 of the garment manufacturer to prevent users who have not purchased the image from using the image without authorization.

[0118] Create a first layer with the above identifier at the top of the image for the first real use; create the first layer containing the identifier (string) at the top of the first real image.

[0119] Calculate the first length, which is the average font thickness of the identifier indicated in the first layer; generate the first length, which is the average font thickness of the identifier (string). In other words, the thicker the letter, the larger the first length calculated.

[0120] Create a second layer to copy the pixels below the opaque pixels indicated in the first layer from the image for the first real use; in the first layer, a portion of the first real image located at the bottom of the first layer is copied and created as the second layer of the second layer.

[0121] Move the 2-1 layer to the left by the first length; the 2-1 layer is parallel to the left side, distorting a portion of the image in the first real use.

[0122] Create a second layer to copy the pixels below the opaque pixels indicated in the first real image, i.e. the second layer; copy a portion of the first real actual image below the portion where the 2-1 layer is located, and then create the 2-2 layer.

[0123] Move the 2-2 layer to the right by twice the length of the first layer; the 2-2 layer is parallel to the right side, distorting the image portion of the first thread.

[0124] Create a second layer to a third layer to copy the pixels below the opaque pixels indicated in the first layer from the image for the first real use; in the middle, a portion of the image for the first real use located at the bottom of the first layer is copied and created as the second layer and the third layer.

[0125] Mosaic pixelization of the 2-3 layer; in the middle, the opaque pixels contained in the second layer and the third layer are pixelated as a mosaic to distort the image.

[0126] Delete the first layer; delete the first layer, which is a portion of the identifier (string) originally created.

[0127] Figure 5A A processing example up to this stage is shown.

[0128] Merge the 2-1 layer, the 2-2 layer, and the 2-3 layer to create a 3-1 layer L31; the layers 2-1 to 2-3 are merged and considered as one layer.

[0129] The border area A located outside the first length is converted from the opaque pixels indicated in the 3-1 layer L31 to the first color; in the middle, the outline is extracted from the opaque pixels of the 3-1 layer L31, which is a single layer, and a contour area with a thickness equal to the length of the first color is created from it as a border (band) of the first color.

[0130] The image of the border area A of the first color is generated as shown in Figure 5B

[0131] At this time, it is desirable to designate the first color as a color not included in the image at the time of the first actual use so that it can be clearly distinguished from a part of the existing image at the time of the first actual use.

[0132] A 3-2 layer L32 is created that copies the pixels located at the bottom of the first color indicated in the 3-1 layer L31 from the image for the first actual use; the pixels located at the bottom of the border area A are copied to create the 3-2 layer L32.

[0133] The 3-2 layer L32 is subjected to chroma inversion; in the middle, the color inversion layer 3-2 L32 is converted to convert the color.

[0134] A viewer image is returned in which the 3-2 layer L32 is located at the top, the 3-1 layer L31 is located at the center, and the image for the first actual use is located at the bottom; by adjusting the order of each layer, you can determine which layer is layered and which layer is placed below.

[0135] Figure 5C An example of the work done so far is shown.

[0136] If the user deletes the watermark and then uses the image generation function to superimpose, a distorted image will be produced instead of a normal actual image, so it is possible to prevent a user who has not purchased the first party image from randomly capturing on a web page and then deleting the watermark or pasting a random image to the deleted part of the watermark according to an artificial intelligence image generation model M.

[0137] ​In addition, the actual image is an image captured in a 360-degree direction around the model position M, and the second generation step and the seventh generation stage are respectively: a first capture angle and a first capture time of a frame captured from the actual image; according to the frequency of each angle of the first capture angle, a step of extracting a 2-1 capture angle corresponding to the maximum appearance 1 level and a 2-2 capture angle corresponding to the maximum appearance 2 level; according to the frequency of each angle of the first capture angle, a step of extracting a 2-3 capture angle corresponding to the minimum appearance 1 level and a 2-4 capture angle corresponding to the minimum appearance 2 level; according to the frequency of each hour of the first capture time, a step of extracting a 2-1 capture time corresponding to the maximum appearance 1 level and a 2-2 capture time corresponding to the maximum appearance 2 level; according to the frequency of each angle of the first capture time, a step of extracting a 2-3 capture time corresponding to the minimum appearance 1 level and a 2-4 capture time corresponding to the minimum appearance 2 level; extracting the median of the 2-1 capture angle and the 2-2 capture angle and the first frame corresponding to the median of the 2-1 capture time and the 2-2 capture time; extracting the median of the 2-1 capture angle and the 2-2 capture angle and the second frame corresponding to the median of the 2-3 capture time and the 2-4 capture time; extracting the median of the 2-3 capture angle and the 2-4 capture angle and the third frame corresponding to the median of the 2-1 capture time and the 2-2 capture time; extracting the median of the 2-3 capture angle and the 2-4 capture angle and the fourth frame corresponding to the median of the 2-3 capture time and the 2-4 capture time; adding the third capture angle of the fifth frame to the first capture angle of the fifth frame; adding the third capture time of the fifth frame to the first capture time of the fifth frame; and returning the fifth frame, that is, any one of the first frame to the fourth frame; it can be used to include.

[0138] The actual image can be an image captured from a plurality of cameras C installed in a 360-degree direction around the model position M.

[0139] According to an embodiment, when a part of the actual image is captured, the first capture angle and the first capture time corresponding to the angle C specified for each camera and the recording time of the image are recorded.

[0140] Here, when the image captured from the camera C at the 12 o'clock position is 0 degrees, the image captured from the camera C at the 3 o'clock position can be specified as 90 degrees, and for the image captured from the camera C at the 6 o'clock position, the "capture angle" can be specified as 180 degrees.

[0141] The first capture angle and the first capture time of the previous capture frame are read from the actual video; the previously recorded capture is read.

[0142] According to the frequency of each angle of the first capture angle, the step of extracting 2-1 capture angle corresponding to the most appearance 1 level capture angle and 2-2 corresponding to the most appearance 2 level capture angle; Extract 2-1 capture angle (the most common capture angle in the capture record) and 2-2 capture angle (the next most common capture angle).

[0143] According to the frequency of each angle of the first capture angle, the step of extracting 2-3 capture angle corresponding to the lowest appearance 1 level and 2-4 capture angle corresponding to the lowest appearance 2 level; Extract 2-3 capture angle (the least visible capture angle in the capture record) and 2-4 capture angle (the next least visible capture angle).

[0144] According to the frequency of each hour of the first capture time, the step of extracting 2-1 capture time corresponding to the most appearance 1 level and 2-2 capture time corresponding to the most appearance 2 level; Extract 2-1 capture time (the most common capture time in the capture record) and 2-2 capture time (the next most common capture time).

[0145] According to the frequency of each angle of the first capture angle, the step of extracting 2-3 capture angle corresponding to the lowest appearance 1 level and 2-4 capture angle corresponding to the lowest appearance 2 level; Extract 2-3 capture angle (the least visible capture angle in the capture record) and 2-4 capture angle (the next least visible capture angle).

[0146] Here, the capture angle corresponds to the "composition" of the photo, and the capture time corresponds to the "pose" of the model M (because the model M takes pictures by changing the pose according to the time).

[0147] Extract the median of 2-1 capture angle and 2-2 capture angle, and the first frame corresponding to the median of 2-1 capture time and 2-2 capture time; Capture photos with similar composition and pose.

[0148] If an angle has two medians (for example, the median between 12 o'clock and 5 o'clock is 2 o'clock and 3 o'clock), any one of them is designated as the median, and if there is no median (for example, 12 o'clock and 1 o'clock are consecutive values), the median is the median corresponding to the lower value in the level.

[0149] Extract the median of 2-1 capture angle and 2-2 capture angle, and the second frame corresponding to the median of 2-3 capture time and 2-4 capture time; Take photos with similar composition but different poses.

[0150] Extract the median of 2-3 capture angle and 2-4 capture angle, and the third frame corresponding to the median of 2-1 capture time and 2-2 capture time; Take photos with different compositions but similar poses.

[0151] The median of the 2-3 capture angle and the 2-4 capture angle, and the median of the 2-3 capture time and the 2-4 capture time are extracted; in the original version, the composition and the posture are different from the original photo.

[0152] The third capture angle of the fifth frame is added to the first capture angle; the third capture time of the fifth frame is added to the first capture time; by recording the capture information of this capture in the data, the data as the highest and lowest appearance basis in the subsequent capture process can be updated.

[0153] Return to the fifth frame, any one of the first frame to the fourth frame; return all the extracted frames, including some or any one frame, as the fifth frame.

[0154] The apparatus according to an embodiment includes a processor and a memory. The device according to the embodiment can be the server or the terminal described above. The processor can include at least one of the aforementioned devices by the drawings, or perform at least one of the aforementioned methods by the drawings. The memory can store information related to the above-described method, or can store a program for implementing the method. The memory can be a volatile memory, or a non-volatile memory.

[0155] The processor can run a program and control the device. The program code executed by the processor can be stored in the memory. The device is connected to an external device (for example, a personal computer or a network) through an input / output device (not shown in the drawing), and can exchange data.

[0156] The above-described embodiments can be implemented as hardware components, software components, and / or a combination of hardware components and software components. For example, the devices, methods, and components described in the embodiments can be implemented using one or more general-purpose computers or special-purpose computers, such as processors, controllers, arithmetic logic units (ALUs), digital signal processors, microcomputers, field programmable gate arrays (FPGAs), programmable logic units (PLUs), microprocessors, or any other devices capable of executing and responding to instructions. The processing unit can execute an operating system (OS) and one or more software applications executed on the operating system. The processing unit can also access, store, manipulate, process, and generate data in response to the execution of software. For ease of understanding, the processing unit can be described as being used as a whole, but those having ordinary knowledge in the art can know that the processing unit can include multiple processing elements and / or multiple types of processing elements. For example, one processing unit can include multiple processors or one processor and one controller. In addition, other processing configurations can also be used, such as parallel processors.

[0157] The method according to the embodiments can be implemented in the form of program instructions, which can be executed by various computer means and recorded on a computer readable medium. The computer readable medium can contain program instructions, data files, data structures, etc. individually or in combination. The program commands recorded in the medium can be designed and configured specifically for the embodiments, or they can be known and available to computer software artisans. Examples of computer readable recording media include magnetic media (such as hard disks, floppy disks, and magnetic tapes), optical media (such as CD-ROM and DVD), magneto-optical media (such as floptical disks), and hardware devices specially configured to store and execute program commands, such as ROM, RAM, flash memory, etc. Examples of program instructions include machine code, such as machine code generated by a compiler, and high-level language code, such as can be executed by a computer using an interpreter. The hardware device can be configured to operate as one or more software modules to perform the operations of the embodiments, or vice versa.

[0158] The software can include one or more combinations of computer programs, codes, instructions, or the like, and can configure the processing unit to run as needed, or can independently or jointly command the processing unit. The software and / or data can be permanently or temporarily embodied in any type of machine, component, physical device, virtual device, computer storage medium or device, or transmitted signal wave, so as to be interpreted by the processing unit or to provide instructions or data to the processing unit. The software is distributed on a networked computer system, which can be stored or executed in a distributed manner. The software and data can be stored on one or more computer readable recording media.

[0159] Although the above-described embodiments have been described through limited drawings, those having ordinary knowledge in the art can apply various technical modifications and modifications on the basis of the above. For example, if the described technology is executed in a different order from the described method, and / or if the components of the described system, structure, device, circuit, etc. are combined or combined in a different way from the described method, or are replaced or replaced by other components or equivalents, appropriate results can be obtained.

[0160] Therefore, other implementations, other embodiments, and embodiments equivalent to the patent claims also belong to the scope of the following claims.

Claims

1. A method for providing a service of a clothing image diversification and user participation clothing image transaction platform based on video processing, wherein, Comprising: In a method of providing an image processing-based clothing image diversification and user-participating clothing image transaction platform service, a device including a processor, a memory, a communication module, and a non-transitory storage medium is provided, and the method is performed by the processor by executing a program stored in the non-transitory storage medium; receiving an actual image of a first clothing item from a model terminal; creating a first real-time image based on the actual image; uploading the first actual image to an image transaction page; receiving a request to purchase the image from a clothing supplier terminal according to the image transaction page; transmitting a second actual image corresponding to the image purchase request to a clothing manufacturer terminal; and deleting the second actual image from the image transaction page. ​

Images