Method and system for generating three-dimensional object

Abstract

A method for generating a three-dimensional object in a three-dimensional object generation system according to the present invention may comprise the steps of: receiving a selection of any one of a 3D asset generation mode or a 3D character generation mode from a user terminal via an interface provided in the system; and providing, to the user terminal, a generation page corresponding to the selected generation mode.

Description

Method and system for creating three-dimensional objects

[0001] The present invention relates to a method and system for generating a three-dimensional object based on an image.

[0002] Recently, interest has been growing in three-dimensional virtual spaces where social, economic, and cultural activities take place, just like in the real world. These virtual spaces can be implemented not only through devices like computers, laptops, and smartphones, but also through wearable devices, utilizing virtual reality (VR), augmented reality (AR), mixed reality (MR), and extended reality (XR).

[0003] In particular, virtual spaces are being developed in a wide range of fields, from autonomous driving to tourism and gaming, and interest is growing in how to create virtual spaces that resemble real-life spaces to provide users with experiences similar to real life.

[0004] In this regard, the role of virtual 3D objects in visual media such as games, movies, and animations has been gradually increasing, and the value of 3D objects that can focus the attention of consumers is also gradually expanding.

[0005] Meanwhile, the present invention was derived as part of the national research and development project, 'Development of Universal Fashion Creation Platform Technology for Avatar Individuality Expression' (Project Unique Number: 1375027535, Project Number: 00228331, Ministry Name: Ministry of Culture, Sports and Tourism, Project Management (Specialized) Agency Name: Korea Creative Content Agency, Research Project Name: Metaverse Content IP Construction Research and Development, Research Project Name: Development of Universal Fashion Creation Platform Technology for Avatar Individuality Expression, Contribution Rate: 1 / 1, Project Performing Agency Name: Kai Co., Ltd., Research Period: 2023.04.01 ~ 2023.12.31).

[0006] The present invention relates to a method and system for generating a three-dimensional object based on artificial intelligence.

[0007] More specifically, the present invention relates to a method and system for providing an integrated interface for creating a three-dimensional object using an image.

[0008] Furthermore, the present invention relates to a method and system that can create a three-dimensional object using a creation technique desired by a user.

[0009] In order to solve the problem discussed above, a method for generating a 3D object in a 3D object generation system according to the present invention may include a step of selecting one of a 3D asset generation mode and a 3D character generation mode from a user terminal through an interface provided by the system, a step of providing a generation page corresponding to the selected generation mode to the user terminal, a step of selecting one of a plurality of detailed generation models dependent on the selected generation mode from the user terminal on the generation page, a step of specifically receiving an input image from the user terminal, and a step of generating a 3D object according to the selected detailed generation model using a target object included in the input image.

[0010] Furthermore, the system may further include a step of generating a file name of the input image using a point in time specified as the input image, and storing the input image in a database of the system with the generated file name.

[0011] Furthermore, the step of generating the three-dimensional object includes a step of generating a file name of the three-dimensional object, and the file name of the three-dimensional object may include at least a portion of unique information of a specific detailed generation model that generated the three-dimensional object among the plurality of detailed generation models and the file name of the input image.

[0012] Furthermore, the 3D character creation mode includes a first detailed creation model and a second detailed creation model, and the first detailed creation model and the second detailed creation model may differ in at least one of a creation speed and a creation quality for the 3D object.

[0013] Furthermore, the 3D character creation mode includes at least one of a first detailed creation model for creating a 2D stylization for the target object, a second detailed creation model for creating a multi-viewpoint view, and a third detailed creation model for performing 3D model deformation, and the file name of the 3D object may be different depending on which creation model among the first detailed creation model, the second detailed creation model, and the third detailed creation model the 3D object was created using in the 3D character creation mode.

[0014] Furthermore, when the 3D asset creation mode is selected from the user terminal, a creation page of the 3D asset creation mode is provided to the user terminal, and one area of ​​the creation page of the 3D asset creation mode includes an object selection area including at least one 3D character created in the 3D character creation mode, and an extended 3D object can be created using the 3D object and a specific 3D character selected in the object selection area.

[0015] Furthermore, the file name of the 3D object further includes user account information logged into the user terminal, and at least one 3D character including the user account information as a file name among a plurality of 3D characters stored in the system can be loaded in the object selection area.

[0016] Furthermore, the extended 3D object includes the 3D object and the specific 3D character, and in the extended 3D object, a relative positional relationship between the 3D object and the specific 3D character can be specified based on a user input received from the user terminal for the creation page of the 3D asset creation mode.

[0017] Meanwhile, a three-dimensional object generation system according to the present invention includes a communication unit for selecting one of a 3D character generation mode and a 3D character generation mode from a user terminal through an interface provided by the system, and a control unit for providing a generation page corresponding to the selected generation mode to the user terminal, wherein the control unit can select one of a plurality of detailed generation models dependent on the selected generation mode from the user terminal on the generation page, receive an input image from the user terminal, and generate a three-dimensional object according to the selected detailed generation model using a target object included in the input image.

[0018] Meanwhile, a program according to the present invention is a program executed by one or more processes in an electronic device and stored in a computer-readable recording medium, and may include commands for performing a step of selecting one of a 3D character creation mode and a 3D character creation mode from a user terminal through an interface provided in a system, a step of providing a creation page corresponding to the selected creation mode to the user terminal, a step of selecting one of a plurality of detailed creation models dependent on the selected creation mode from the creation page, a step of receiving an input image from the user terminal, and a step of generating a 3D object according to the selected detailed creation model using a target object included in the input image.

[0019] The method and system for generating a 3D object according to the present invention can provide an integrated interface that allows a user terminal to select one of a 3D asset generation mode and a 3D character generation mode through an interface provided by the system, and provides a generation page corresponding to the selected generation mode to the user terminal. Through this, the user can easily access the integrated interface provided by the present invention without installing a separate program, and can generate 3D assets and stylized 3D characters through the integrated interface.

[0020] Furthermore, the method and system for generating a 3D object according to the present invention can receive, from the user terminal, one of a plurality of detailed generation models dependent on the selected generation mode on the generation page, and, after receiving an input image from the user terminal, generate a 3D object according to the selected detailed generation model using a target object included in the input image. The user can generate a 3D object by selecting a desired detailed generation model from among a plurality of detailed generation models, and even if the user does not have the ability to create a 3D object, the user can easily and freely generate a 3D object through the integrated interface of the present invention.

[0021] Furthermore, the method and system for generating a 3D object according to the present invention can generate a file name of the input image using a point in time specified as the input image, and can generate a file name of the 3D object using unique information of a specific detailed generation model that generated the 3D object and the file name of the input image. Through such name rule normalization, it is possible to generate multiple 3D graphic objects simultaneously by maximizing the performance of the memory of the graphics processing unit (GPU).

[0022] Figure 1 is a conceptual diagram for explaining a three-dimensional object creation system according to the present invention.

[0023] Figure 2 is a conceptual diagram for explaining name rule normalization in the present invention.

[0024] Figure 3 is a flowchart for explaining a method for creating a three-dimensional object according to the present invention.

[0025] FIGS. 4A and 4B are conceptual diagrams for explaining a method of creating a three-dimensional object in a 3D asset creation mode according to the present invention.

[0026] FIGS. 5a, 5b and 5c are conceptual diagrams for explaining a method of creating a three-dimensional object in a 3D character creation mode according to the present invention.

[0027] Figures 6 and 7 are conceptual diagrams for explaining a method for generating an extended three-dimensional object in the present invention.

[0028] Hereinafter, embodiments disclosed in this specification will be described in detail with reference to the attached drawings. Regardless of the drawing numbers, identical or similar components will be given the same reference numbers, and redundant descriptions thereof will be omitted. The suffixes "module" and "part" used for components in the following description are assigned or used interchangeably only for the convenience of writing the specification, and do not in themselves have distinct meanings or roles. In addition, when describing the embodiments disclosed in this specification, if it is determined that a specific description of a related known technology may obscure the gist of the embodiments disclosed in this specification, a detailed description thereof will be omitted. In addition, the attached drawings are only intended to facilitate easy understanding of the embodiments disclosed in this specification, and the technical ideas disclosed in this specification are not limited by the attached drawings, and should be understood to include all modifications, equivalents, and substitutes included in the spirit and technical scope of the present invention.

[0029] Terms that include ordinal numbers, such as first, second, etc., may be used to describe various components, but the components are not limited by these terms. These terms are used solely to distinguish one component from another.

[0030] When a component is referred to as being "connected" or "connected" to another component, it should be understood that it may be directly connected or connected to that other component, but that there may be other components intervening. Conversely, when a component is referred to as being "directly connected" or "connected" to another component, it should be understood that there are no other components intervening.

[0031] Singular expressions include plural expressions unless the context clearly indicates otherwise.

[0032] In this application, terms such as “include” or “have” are intended to specify the presence of a feature, number, step, operation, component, part or combination thereof described in the specification, but should be understood not to exclude in advance the possibility of the presence or addition of one or more other features, numbers, steps, operations, components, parts or combinations thereof.

[0033] The present invention provides an integrated interface for creating three-dimensional asset objects and three-dimensional character objects, enabling multiple users to use the integrated interface without installing any separate programs, and relates to a three-dimensional object creation method and system capable of simultaneously creating a plurality of three-dimensional graphic objects by utilizing the maximum performance of the memory of a graphics processing unit (GPU) through name rule normalization. The “dimensional object creation system” described in the present invention may also be referred to as a “dimensional object creation server” or a “dimensional object creation platform.”

[0034] Hereinafter, a method and system for generating a three-dimensional object according to the present invention will be described in detail with reference to the attached drawings. Fig. 1 is a conceptual diagram for explaining a three-dimensional object generation system according to the present invention. Fig. 2 is a conceptual diagram for explaining name rule normalization according to the present invention, Fig. 3 is a flowchart for explaining a three-dimensional object generation method according to the present invention, Figs. 4a and 4b are conceptual diagrams for explaining a method for generating a three-dimensional object in a 3D asset generation mode according to the present invention, Figs. 5a, 5b and 5c are conceptual diagrams for explaining a method for generating a three-dimensional object in a 3D character generation mode according to the present invention, and Figs. 6 and 7 are conceptual diagrams for explaining a method for generating an extended three-dimensional object according to the present invention.

[0035] As illustrated in FIG. 1, the three-dimensional object system (100) according to the present invention may be configured to include at least one of a communication unit (110), a storage unit (120), and a control unit (130).

[0036] The communication unit (110) can be connected to an external device via a wireless or wired network. Accordingly, the communication unit (110) can receive information required for the overall operation of the system (100) from the external device.

[0037] The communication unit (110) can provide the user terminal (10) with an interface (or service page, also named 200) that provides a 3D object creation function to the user terminal (10). The communication unit (110) can receive various information related to 3D object creation (e.g., an image, a user selection for a 3D object creation model, etc.) from the user terminal (10) through the interface (200). In addition, the communication unit (110) can provide the created 3D object to the user terminal (10) through the interface (200).

[0038] The interface (200) may correspond to a web page or a page (or screen) of an application, and there is no limitation on the type of page as long as it is a page that can be provided to a user terminal (10).

[0039] The user terminal (10) is an electronic device capable of outputting an interface (200), and may include, for example, at least one of a mobile phone, a tablet PC, a head-mounted display, XR glasses, a computer, a laptop, a digital broadcasting terminal, a PDA (Personal Digital Assistant), a PMP (Portable Multimedia Player), and a KIOSK.

[0040] The storage unit (120) can store data and commands required for the overall operation of the system (100) according to the present invention. For example, the storage unit (120) can store commands required for creating a three-dimensional object. In the present invention, the storage unit (120) can be used interchangeably with a database (DB), and the database can also be given the same drawing symbol “” as the storage unit.

[0041] In the storage unit (120), an input image specified by the user terminal (10) and a 3D object created using the input image can be stored.

[0042] As illustrated in (a) of FIG. 2, in the storage unit (120), an input image (Input) used for creating a 3D object and a 3D object (Ouput) created in the present invention can be normalized and stored according to a predefined name rule.

[0043] Here, the input image and 3D object can be named as document, content, file, etc., and the names of the input image and 3D object can be understood as data name, document name, file name, folder name, etc. In the present invention, for the convenience of explanation, the names of the input image and 3D object are named as “file name” and explained.

[0044] The file name of the input image (Input) stored in the storage unit (120) may be normalized according to the rule of “{image_Time_ID}.jpg”. More specifically, in the present invention, the input image may be normalized based on at least one of the time information (Time) and user account information (ID) specified as the input image.

[0045] The input image may be specified in various ways. For example, the input image may be specified based on a user input for uploading the input image to the system from a user terminal, receiving the input image from the user terminal, or selecting at least one of a plurality of sample images already registered in the system as the input image. Accordingly, in the present invention, the time information may be determined based on various criteria, such as the time at which the input image is uploaded to the system, the time at which the input image is received from the user terminal, or the time at which a specific sample image is selected from the user terminal.

[0046] For example, let's assume that a specific image is received as an input image from a user terminal at 15:30:25 on November 27, 2023. The input image may be stored in the storage unit (120) with the file name {image_20231127153025}.jpg. In this case, the time information may serve as user information that identifies the user.

[0047] For another example, when a specific input image is received as an input image from a “”user account, the input image may be stored in the storage unit (120) with the file name “{image_20231127153025_KaiStudio}.jpg”.

[0048] In the present invention, there is no limitation on the unit of time information, and in order to distinguish input images received at close times, time information in small units such as 1 / 1000th of a second or 1 / 10,000th of a second can be used.

[0049] Furthermore, the file name of the 3D object (Output) stored in the storage unit (120) may be normalized and exist according to the rule of “{model_name} / {image_Time_ID}.png”. More specifically, in the present invention, the 3D object may be normalized based on at least one of the unique information of the generation model used to generate the 3D object (e.g., the generation model name, identification number, etc.) and the file name of the input image.

[0050] For example, if a 3D object is generated based on a fast but low-quality 3D generation model among multiple generation models, the 3D object may be stored in the storage unit (120) with a file name of “{LowQual_Fast} / {image_20231127153025_KaiStudio}.png”. For another example, if a 3D object is generated based on a slow but high-quality (1000 epochs) 3D generation model for the same input image, the 3D object may be stored in the storage unit (120) with a file name of “{HighQual_1000} / {image_20231127153025_KaiStudio}.png”.

[0051] Furthermore, when an extended 3D object is created by combining (or synthesizing) multiple 3D objects (Output 1 and Output 2), the file name of the corresponding 3D object (Output 3) can be specified based on the file names of the multiple 3D objects. For example, as illustrated in (b) of FIG. 2, if the file name of the first object (Output 1) is “{model_name1} / {image_Time1_ID}.png” and the file name of the first object (Output 2) is “{model_name2} / {image_Time2_ID}.png”, the extended 3D object (Output 3) can be stored in the storage unit (120) with the file name “{merge} / {model_name1} / {image_Time1_ID} / {model_name2} / {image_Time2_ID}.png”.

[0052] In this way, in the present invention, through file name rule normalization based on at least one of time information, user account, and creation model, multiple three-dimensional graphic objects can be created simultaneously within the range permitted by the memory of a graphic processing unit (GPU).

[0053] Meanwhile, the control unit (130) can control the overall operation of the system (100) according to the present invention. The control unit (130) can generate a three-dimensional object corresponding to an input image by using a specific generation model among a plurality of generation models (or detailed generation models) based on user input.

[0054] The control unit (130) can create a three-dimensional object by using a plurality of three-dimensional detailed creation models (141, 142, 151, 152, 153) corresponding to one configuration of the present invention, or by using a creation model provided from an external server. When using a creation model from an external server, the control unit (130) can create a three-dimensional object through communication with the external server. For convenience of explanation, the system (100) according to the present invention will be described below as including a plurality of three-dimensional detailed creation models (141, 142, 151, 152, 153).

[0055] A plurality of three-dimensional detailed generation models (141, 142, 151, 152, 153) can be classified into a first type generation model (140) and a second type generation model (150) depending on the three-dimensional object generation method.

[0056] The first type of generative model (140) can be understood as a model that generates a three-dimensional mesh corresponding to a single input image and generates a three-dimensional object using the three-dimensional mesh. In the present invention, the first type of generative model (140) may include a plurality of detailed generative models (141, 142) that differ in at least one of the generation speed and the generation quality for the three-dimensional object.

[0057] The first detailed generation model (also called “first 3D generation model” or “fast generation model”, 141) may be a model whose generation speed for a 3D object is relatively faster than that of the second detailed generation model (also called “second 3D generation model” or “high-quality generation model”, 142), but whose generation quality for a 3D object is relatively lower than that of the second detailed generation model (142).

[0058] The second detailed generation model (142) may correspond to a model whose generation speed for a 3D object is relatively slower than that of the first detailed generation model (141), but whose generation quality for a 3D object is relatively higher than that of the first detailed generation model (141).

[0059] The second type of generative model (150) is a generative model that generates a stylized 3D character. It can be understood as a generative model that first generates a stylized image using an input image including a person and a reference image, and then generates a 3D character based on the generated stylized image.

[0060] In the present invention, the second type of generation model (150) may include at least one of a first detailed generation model (which may be named a “styling model”, 151) that generates a 2D stylization for a target object, a second detailed generation model (which may be named a “multi-view view generation model”, 152) that generates a multi-view view, and a third detailed generation model (which may be named a “deformation model”, 153) that performs a 3D model deformation.

[0061] The control unit (130) can create a 3D object from one of a plurality of 3D detailed creation models (141, 142, 151, 152, 153) based on a user's selection, normalize a file name for the 3D object, and store the 3D object in a database (120) with the corresponding file name. Hereinafter, a method for creating a 3D object will be described in more detail.

[0062] In the present invention, a process of selecting one of a 3D asset creation mode and a 3D character creation mode from a user terminal through an interface provided by the system may be performed (S310, see FIG. 3). In addition, in the present invention, a process of providing a creation page corresponding to the selected creation mode to the user terminal may be performed (S320, see FIG. 3).

[0063] The control unit (130) may provide an interface (200) including a creation mode selection function to the user terminal (10). For example, as illustrated in FIG. 1, the interface (200) may include a first tab (210a) corresponding to a 3D asset creation mode and a second tab (220a) corresponding to a 3D character creation mode. The first tab (210a) may be matched with a creation page (210) of the 3D asset creation mode, and the second tab (220a) may be matched with a creation page (220) of the 3D character creation mode.

[0064] The control unit (130) may provide a creation page (either 210 or 220) matching the selected tab on the interface (200) based on whether one of the first tab (210a) and the second tab (220a) is selected. More specifically, the control unit (130) may provide a creation page (210) of a 3D asset creation mode on the interface based on whether the first tab (210a) is selected, and may provide a creation page (220) of a 3D character creation mode on the interface based on whether the second tab (220a) is selected. In this case, one of the first tab (210a) and the second tab (220b) is set to a default value, and the control unit (130) may provide a creation page matching the tab set to the default value on the interface based on an interface output request. For example, if the first tab (210a) is set to the default value, the control unit (130) can provide a creation page (210) of a 3D asset creation mode matching the first tab (210a) on the interface based on an interface (200) output request from the user terminal (10).

[0065] In the present invention, a process of selecting one of a plurality of detailed generation models dependent on a selected generation mode of a generation page may be performed from a user terminal (S330, see FIG. 2).

[0066] The control unit (130) can select any one of a plurality of detailed creation models dependent on each creation mode through the creation page (210) of the 3D asset creation mode or the creation page (220) of the 3D character creation mode.

[0067] The control unit (130) may provide a plurality of detailed creation mode items (211 to 213, 221 to 223) dependent on each creation mode on the creation page (210) of the 3D asset creation mode or the creation page (220) of the 3D character creation mode. When a specific item among the items provided on the creation page is selected, the control unit (130) may specify a detailed creation model corresponding to the selected item as a detailed creation model to be used for 3D object creation.

[0068] More specifically, as illustrated in FIG. 1, the control unit (130) may provide, on the creation page (210) of the 3D asset creation mode, a first item (fast but low quality, 211) corresponding to the first 3D creation model (141), a second item (slow but high quality (1000 epochs), 212) corresponding to the second 3D creation model (142), and a third item (slow but high quality (3000 epochs), 213) corresponding to the third 3D creation model (not shown), based on the selection of the first tab (210a). When any one of the first to third items (211 to 213) is selected, the control unit (130) may generate a 3D object using the detailed creation model corresponding to the selected item.

[0069] Furthermore, the control unit (130) may provide, on the creation page (220) of the 3D character creation mode, a first item (2D stylization, 221) corresponding to the 2D stylization model (151), a second item (multi-viewpoint view creation, 222) corresponding to the multi-viewpoint view creation model (152), and a third item (3D model deformation, 223) corresponding to the 3D deform model (153) based on the selection of the second tab (220a). When any one of the first to third items (221 to 223) is selected, the control unit (130) may create a 3D object using the detailed creation model corresponding to the selected item.

[0070] Meanwhile, as described above, the control unit (130) can generate a 3D object based on a specific detailed generation model, and normalize the file name of the 3D object based on the unique information of the specific detailed generation model and the file name of the input image. In this case, the file name of the input image can be normalized based on the time information at which the input image is specified. Therefore, in the present invention, by normalizing the file names of the input image and the 3D object based on the time at which the user specifies the input image and the detailed generation model selected by the user, a plurality of 3D objects can be distinguished and generated simultaneously.

[0071] As illustrated in FIG. 4A, based on which one of a plurality of detailed generation models dependent on a 3D asset generation mode is selected, the control unit (130) can provide a generation page (400) of a 3D asset generation mode corresponding to a specific detailed generation model to the user terminal (10). The control unit (130) can receive various commands for generation and storage of a specific 3D object of an input image through the generation page (400).

[0072] More specifically, the creation page (400) of the 3D asset creation mode may include at least one of a first area (410) that receives an input image, a second area (also referred to as an “object viewer area”, 420) that provides a created 3D object, and a third area (430) that receives various commands related to creating a 3D object.

[0073] In the present invention, a process of receiving a specific input image from a user terminal may be performed (S340, see FIG. 3). In addition, in the present invention, a process of generating a 3D object according to a selected detailed generation model may be performed using a target object included in the input image (S350, see FIG. 3).

[0074] The control unit (130) can specify an input image (411) through the first area (410). There may be various methods for specifying the input image (411). For example, the control unit (130) can specify an image as an input image based on the fact that a specific image is uploaded to the first area (410). As another example, the control unit (130) can provide a plurality of sample images registered in the system to the first area (410) and specify an input image based on which one of the plurality of sample images is selected.

[0075] In the present invention, a three-dimensional object can be created according to a selected detailed creation model using a target object included in an input image.

[0076] When the input image (411) is specified, the control unit (130) can specify (or detect) a target object (e.g., a single-seater sofa, 411a) that is the target of three-dimensional object generation from the input image (411).

[0077] The control unit (130) can generate a three-dimensional mesh corresponding to the detected target object (411a) based on a specific detailed generation model. In addition, the control unit (130) can apply the three-dimensional mesh to the target object (411a) to generate a three-dimensional object corresponding to the target object.

[0078] For example, the control unit (130) can search for a mesh model corresponding to a specific detailed creation model and a target object (411a) from the database (120). The database (120) can include a plurality of different mesh models according to categories for each of a plurality of detailed creation models. The control unit (130) can generate a 3D mesh by specifying (obtaining or referencing) mesh model information that is most similar to the target object (411a) from the database (120) corresponding to a specific detailed creation model.

[0079] For another example, feature point information corresponding to a target object may be matched and stored in advance for each of a plurality of mesh models included in the database (120). The control unit (130) may extract feature points from the target object (411a), and, among the feature point information corresponding to the feature points extracted from the target object (411a) among the feature point information matched to each of the plurality of mesh models, may generate a mesh model corresponding to the target object (411a).

[0080] Meanwhile, the above-described method is one embodiment of a method for generating a three-dimensional object, and the control unit (130) can generate a three-dimensional object corresponding to the target object (411a) in various ways according to a selected specific detailed generation model.

[0081] As illustrated in FIG. 4b, the control unit (130) can generate a three-dimensional object (421) corresponding to the target object (411a) based on a selected specific detailed generation model, and provide the generated three-dimensional object (421) through the second area (or viewer area, 430).

[0082] The third region (430) may include graphic objects corresponding to various functions related to 3D object creation. For example, the third region (430) may include graphic objects corresponding to a detailed creation model change function, an optimization loop count change function (e.g., 1000 epochs or 3000 epochs), and a 3D object creation function.

[0083] Meanwhile, the control unit (130) can generate a file name for the input image (411) and the 3D object (421) generated based on the input image (411), and store the input image (411) and the 3D object (421) in the database (120) with the generated file name.

[0084] As described above, the control unit (130) can generate a file name of the input image (411) based on at least one of the time information and user account information specified as the input image (411). For example, if the input image (411) is specified from the user terminal (10) at 15:30:25 on November 27, 2023, the file name of the input image (411) can be generated as "{image_20231127153025}.jpg".

[0085] In addition, the control unit (130) can generate a file name of a 3D object (421) based on unique information of a specific detailed creation model of the file name of the input image (411). In this case, the file name of the 3D object can be generated differently based on whether the specific detailed model is different.

[0086] For example, when a 3D object (421) is generated based on the first 3D creation model (411), the control unit (130) can generate the file name of the 3D object (412) as “{LowQual_Fast} / {image_ 20231127153025_KaiStudio}.png”. For another example, when a 3D object (421) is generated based on the second 3D creation model (412) based on 1000 epochs, the control unit (130) can generate the file name of the 3D object (412) as “{HighQual_1000} / {{image_20231127153025_KaiStudio}.png”. For another example, when a 3D object (421) is created based on a third 3D generation model (not shown) based on 3000 epochs, the control unit (130) can create a file name of the 3D object (412) as “{HighQual_3000} / {{image_20231127153025_KaiStudio}.png”.

[0087] Meanwhile, as illustrated in FIG. 5A, based on which one of a plurality of detailed creation models dependent on a 3D character creation mode is selected, the control unit (130) can provide a creation page (500) of a 3D character creation mode corresponding to a specific detailed creation model to the user terminal (10). The control unit (130) can receive a command for creation and storage of a specific 3D object of an input image and a reference image (or reference image) through the creation page (500).

[0088] The creation page (500) of the 3D character creation mode may include at least one of a first area (510) for receiving an input image, a second area (520) for receiving a reference image, a third area (530) for providing a stylized image, a fourth area (also referred to as a “viewer area”, 540) for providing a generated 3D object, and a fifth area (550) for receiving various commands related to 3D object creation.

[0089] The control unit (130) can receive an input image (511) through the first region (510) and detect a target object (511a) from the specified input image (511).

[0090] In this case, the control unit (130) can detect a specific image object corresponding to a person (or person) among multiple image objects included in the input image (511) as a target object (511a). In the 3D asset creation mode, the control unit (130) detects the target object without distinguishing the type of the object, but in the 3D character creation mode, it can detect an object of the person type as the target object.

[0091] The control unit (130) can receive a reference image (521) through the second area (520). The control unit (130) can control the detailed generation model specified using the reference image (521) to generate a character image (531) of a person target object (511a) corresponding to the reference image (521). As illustrated in FIG. 5b, the control unit (130) can provide at least one character image (531) for the person target object (511a) through the third area (530).

[0092] Here, the reference image (521) is an image used to create a character image (531) of a person target object (511a), and the control unit (130) can receive the reference image from the user terminal (10) or specify the reference image (521) based on the selection of one of a plurality of sample reference images pre-registered in the system.

[0093] The control unit (130) can generate a character image (531) for a person target object (511a) using a specific detailed generation model so as to correspond to the style (or drawing style) of the reference image (521). That is, the control unit (130) can generate a character image (531) by editing the person target object (511a) in the style of the reference image (521).

[0094] In this case, the control unit (130) can generate a character image of a single viewpoint (e.g., a character image of the front, 531a) or a character image of multiple viewpoints (e.g., “viewpoints”) (front, back, left, right character images, 531a, 531b, 531c, 531d). The viewpoint of the character image (531) can be determined based on the user's selection. For example, as illustrated in FIG. 5b, the control unit (130) can generate a character image of a single viewpoint when a single viewpoint button (single, 551) is selected, and can generate a character image of multiple viewpoints (531a, 531b, 531c, 531d) when a multiple viewpoint button (multiple, 552) is selected.

[0095] The control unit (130) can generate a three-dimensional object (also called “dimensional character object”, “character”, “character object” or “avatar”, 541) corresponding to a person target object (511a) according to a specific detailed generation model based on at least one character image (531).

[0096] For example, the control unit (130) can perform volume optimization based on a character image (531) to generate a three-dimensional character model, and build a bone skeleton of the three-dimensional character model through rigging of the three-dimensional character model. Then, the control unit (130) can connect the bone skeleton to the three-dimensional character model through skinning to generate a final three-dimensional character object (541).

[0097] As illustrated in FIG. 5c, the control unit (130) can provide a three-dimensional character object (541) through the fourth area (or viewer area, 540).

[0098] The fifth area (550) may include graphic objects corresponding to various functions related to the creation of three-dimensional character objects. For example, the third area may include graphic objects corresponding to a detailed creation model change function, a reference image selection function, a character image viewpoint selection function, and a three-dimensional character object creation function.

[0099] Meanwhile, the control unit (130) can generate a file name of an input image (511) and a 3D character object (541) generated based on the input image (511), and store the input image (511) and the 3D character object (541) in a database (120) according to the generated file name.

[0100] As described above, the control unit (130) can generate a file name of the input image (511) based on at least one of the time information and the user account information specified as the input image (511). For example, if the input image (511) is specified from the user terminal (10) at 15:30:25 on November 27, 2023, the file name of the input image (511) can be generated as {image_20231127153025}.jpg. In addition, the control unit (130) can generate a file name of the 3D character object (541) based on the unique information of the file name-specific detailed generation model of the input image (511). In this case, the file name of the 3D object can be generated differently based on the fact that the specific detailed model is different.

[0101] More specifically, the 3D character generation mode may include at least one of a first detailed generation model (151) that generates 2D stylization for a person target object (511a), a second detailed generation model (152) that generates a multi-view view, and a third detailed generation model (153) that performs 3D model deformation. The control unit (130) may generate a different file name for the 3D character object (541) depending on which generation model among the first detailed model to the third detailed model (151 to 153) the 3D character object (541) was generated with. For example, if the 3D character object (541) is generated based on a detailed generation model that uses a single-view viewpoint, the control unit (130) may generate the file name of the 3D character object (541) as “{character_singleview} / {image_20231127153025_ KaiStudio}.png”.

[0102] Meanwhile, the control unit (130) can register a 3D object as an asset of a user account by matching the 3D object created through the interface to a user account and storing the same in a database (120).

[0103] In this case, the control unit (130) can set a name for the 3D object that is different from the file name of the 3D object. For example, if the 3D object is an avatar, the control unit (130) can set a nickname for the avatar, and link the set nickname with the 3D avatar object and store them in the database (120).

[0104] The control unit (130) can store input images and 3D objects specified in the system (100) in the database (120) based on a user account. That is, the control unit (130) can distinguish and store data used in the present invention based on a user account (e.g., ID).

[0105] Furthermore, the control unit (130) can provide a 3D object creation service to create an extended 3D object using a 3D object matched to a user account.

[0106] An extended 3D object can be understood as a new 3D object created by combining (or synthesizing) multiple 3D objects created by a specific user account.

[0107] The control unit (130) can create a new type of extended 3D object using a 3D object created in the 3D asset creation mode and a 3D object created in the 3D character creation mode. In the following, in order to avoid confusion in terminology, a 3D object created in the 3D asset creation mode will be named a 3D asset object (also referred to as “asset”, “asset object”, or “first type 3D object”), and a 3D object created in the 3D character creation mode will be named a 3D character object (“character”, “character object”, or “second type 3D object”). In addition, a 3D object created using a 3D asset object and a 3D character object will be named an extended 3D object (or a third type 3D object).

[0108] When one of the 3D asset creation mode and the 3D character creation mode is selected, the control unit (130) can provide a 3D object created in another creation mode in an area of ​​the creation page corresponding to the selected mode. The control unit (130) can create an extended 3D object using the 3D object corresponding to the selected mode and the 3D object created in the other creation mode.

[0109] As illustrated in FIG. 6, when the 3D asset creation mode (or the first tab, 210a) is selected from the user terminal, the control unit (130) may provide a creation page (600) of the 3D asset creation mode to the user terminal. The creation page (600) of the 3D asset creation mode may include at least one of an object viewer area (610) in which a 3D asset object (611) created in the 3D asset creation mode is provided, an object selection area (hereinafter, referred to as a “basket area” to avoid confusion of terms, 620) in which at least one 3D character object (621, 622) created in the 3D character creation mode is provided, and an extended object viewer area (630) in which an extended 3D object is provided.

[0110] The control unit (130) can load a 3D character object (621, 622) matching a user account and display it in the basket area (620) of the creation page (600) corresponding to the 3D asset creation mode.

[0111] In this case, the 3D character object (621, 622) matched to the user account can be understood as a 3D character object that includes, as a file name, information on a specific user account logged into the user terminal (10) among multiple 3D character objects stored in the system. That is, a 3D character object created in the 3D character creation mode by a specific user account can be displayed in the basket area (620).

[0112] When at least one specific 3D character object (621) is selected by a user from among a plurality of 3D character objects (621, 622) included in a basket area (620), the control unit (130) can create an extended 3D object (631) using the selected specific 3D character object (621) and a 3D object (3D asset object, 611) created in a 3D asset creation mode. In addition, the control unit (130) can provide the extended 3D object (631) to an extended object viewer area (630).

[0113] The extended 3D object (631) may include a 3D asset object (611) and a selected specific 3D character object (621). In addition, in the extended 3D object (631), a relative positional relationship between the 3D asset object (611) and the specific 3D character object (621) may be specified based on a user input received from a user terminal (10) for a creation page (600) of a 3D asset creation mode.

[0114] More specifically, based on the user input, the positional relationship between the 3D asset object (611) and the 3D character object (621) on the viewer area (610) is specified, and the control unit (130) can specify the positional relationship between the 3D asset object (611) and the specific 3D character object (621) in the extended 3D object (631).

[0115] For example, as illustrated in FIG. 6, let us assume that the 3D asset object (611) is a “sofa.” The control unit (130) may generate an extended 3D object (631) in which the 3D asset object (611) is sitting on the “sofa” based on a user input of dragging a specific 3D character object (621) from the basket area (620) to the viewer area (610) so that the specific 3D character object (621) overlaps with the 3D asset object (611).

[0116] As another example, although not shown, the control unit (130) may generate a three-dimensional extended object of a specific 3D character object (621) sitting on the floor in front of the 3D asset object (611) based on a user input of dragging a specific 3D character object (621) to a rear area of ​​the 3D character object (611) without overlapping the 3D asset object (611).

[0117] Furthermore, the control unit (130) can specify the positional relationship between the 3D asset object (611) and a specific 3D character object (621) in the extended 3D object (631) based on the type of the 3D asset object (611).

[0118] As illustrated in FIG. 6, when a 3D character object (621) and a 3D asset object “sofa” are positioned in the viewer area (610) so as to overlap, the control unit (130) can generate an extended 3D object in which the 3D character object (621) is sitting on the 3D asset object (611) sofa. Although not illustrated, when the category of the 3D asset object is “pencil” and the 3D character object (621) and the 3D asset object “pencil” are positioned in the viewer area (610) so as to overlap, the control unit (130) can generate a 3D extended graphic object in which the 3D character object (621) is holding the pencil 3D asset object in its hand.

[0119] In this way, the control unit (130) can generate an extended 3D object (631) having different visual appearances based on the category of the 3D asset object (611) and the positional relationship between the 3D asset object (611) and the 3D character object (621) specified in the viewer area (610).

[0120] Meanwhile, when an extended 3D object (631) is created, the control unit (130) can create a file name (e.g., “{model_name1} / {image_Time1_ID}.png”) of the extended 3D object (631) based on the file name (e.g., “{model_name1} / {image_Time1_ID}.png”) of the 3D asset object (611) used to create the extended 3D object (631) and the file name (e.g., “{model_name2} / {image_Time2_ID}.png”) of the 3D character object (621), and store the extended 3D object with the created file name in the database (120).

[0121] Furthermore, as illustrated in FIG. 7, when the 3D character creation mode (or the second tab, 220a) is selected from the user terminal, the control unit (130) may provide a creation page (700) of the 3D character creation mode to the user terminal. The creation page (700) of the 3D character creation mode may include at least one of an object viewer area (710) in which a 3D character object (711) created in the 3D character creation mode is provided, an object selection area (hereinafter, also referred to as a “basket area” to avoid confusion of terms, 720) in which at least one 3D asset object (721, 722) created in the 3D asset creation mode is provided, and an extended object viewer area (730) in which an extended 3D object is provided.

[0122] The control unit (130) can load a 3D asset object (721, 722) matching a user account and display it in the basket area (720) of the creation page (700) corresponding to the 3D character creation mode.

[0123] In this case, the 3D asset object (721, 722) matched to the user account can be understood as a 3D asset object that includes, in its file name, information on a specific user account logged into the user terminal (10) among multiple 3D asset objects stored in the system. That is, it can be understood as a 3D asset object created in the 3D asset creation mode according to the present invention by the user account.

[0124] When a specific 3D asset object (722) is selected by a user from among a plurality of 3D asset objects (721, 722) included in a basket area (720), the control unit (130) can create an extended 3D object (731) using the selected specific 3D asset object (722) and a 3D object (3D character object, 711) created in a 3D character creation mode. In addition, the control unit (130) can provide the extended 3D object (731) to an extended object viewer area (730).

[0125] The extended 3D object (731) may include a 3D character object (711) and a selected specific 3D asset object (722). In addition, in the extended 3D object (731), a relative positional relationship between the 3D character object (711) and the specific 3D asset object (722) may be specified based on a user input received from a user terminal (10) for a creation page (700) of a 3D character creation mode.

[0126] More specifically, based on a positional relationship between a 3D character object (711) and a specific 3D asset object (722) on a viewer area (710) based on a user input, the control unit (130) can specify a positional relationship between the 3D character object (711) and the specific 3D asset object (722) in an extended 3D object (731).

[0127] For example, as illustrated in FIG. 7, assume that a 3D asset object (722) is a “ribbon.” The control unit (130) may create an extended 3D object (731) in which a 3D character object (711) is wearing the “ribbon” of the specific 3D asset object (722), based on a user input of dragging a specific 3D asset object (722) from the basket area (720) to the viewer area (710) so that it overlaps with a 3D character object (711).

[0128] For another example, although not shown, the control unit (130) may generate an extended 3D object (731) in which a 3D character object (711) wears a “ribbon” of a specific 3D asset object (722) on its neck, based on the positioning of the specific 3D asset object (722) to overlap a neck area of ​​the 3D character object (711). Alternatively, the control unit (130) may generate an extended 3D object (731) in which a 3D character object (711) wears a “ribbon” of a specific 3D asset object (722) on its head, based on the positioning of the specific 3D asset object (722) to overlap a head area of ​​the 3D character object (711).

[0129] Furthermore, the control unit (130) can specify the positional relationship between a 3D character object (711) and a specific 3D asset object (722) in an extended 3D object (731) based on the type of the 3D asset object (722).

[0130] As illustrated in FIG. 7, based on the fact that the category of the 3D character object (711) is “ribbon,” the control unit (130) can generate an extended 3D object in which the 3D character object (711) is wearing the 3D asset object (722) “ribbon.” In contrast, although not illustrated, if the category of the 3D asset object is “bed,” and the 3D asset object and the 3D character object “bed” are positioned in the viewer area (710) to overlap, the control unit (130) can generate an extended 3D graphic object in which the 3D asset object (712) is lying on the “bed” 3D asset object.

[0131] Meanwhile, when an extended 3D object (731) is created, the control unit (130) can create a file name of the extended 3D object (731) based on the file name of the 3D character object (711) and the file name of the 3D asset object (722) used to create the extended 3D object (731), and store the extended 3D object with the created file name in the database (120).

[0132] The method and system for generating a 3D object according to the present invention can provide an integrated interface that allows a user terminal to select one of a 3D asset generation mode and a 3D character generation mode through an interface provided by the system, and provides a generation page corresponding to the selected generation mode to the user terminal. Through this, the user can easily access the integrated interface provided by the present invention without installing a separate program, and can generate 3D assets and stylized 3D characters through the integrated interface.

[0133] Furthermore, the method and system for generating a 3D object according to the present invention can receive, from the user terminal, one of a plurality of detailed generation models dependent on the selected generation mode on the generation page, and, after receiving an input image from the user terminal, generate a 3D object according to the selected detailed generation model using a target object included in the input image. The user can generate a 3D object by selecting a desired detailed generation model from among a plurality of detailed generation models, and even if the user does not have the ability to create a 3D object, the user can easily and freely generate a 3D object through the integrated interface of the present invention.

[0134] Furthermore, the method and system for generating a 3D object according to the present invention can generate a file name of the input image using a point in time specified as the input image, and can generate a file name of the 3D object using unique information of a specific detailed generation model that generated the 3D object and the file name of the input image. Through such name rule normalization, it is possible to generate multiple 3D graphic objects simultaneously by maximizing the performance of the memory of the graphics processing unit (GPU).

[0135] Meanwhile, the present invention discussed above can be implemented as computer-readable code or instructions on a program-recorded medium. In other words, the present invention can be provided in the form of a program.

[0136] Computer-readable media include any type of storage device that stores data that can be read by a computer system. Examples of computer-readable media include hard disk drives (HDDs), solid-state disk drives (SSDs), silicon disk drives (SDDs), ROMs, RAMs, CD-ROMs, magnetic tapes, floppy disks, and optical data storage devices.

[0137] Furthermore, the computer-readable medium may include a storage device and may be a server or cloud storage device accessible to a user terminal via communication. In this case, the computer may download the program according to the present invention from the server or cloud storage device via wired or wireless communication.

[0138] Furthermore, in the present invention, the computer described above is a user terminal equipped with a processor, i.e., a CPU (Central Processing Unit), and there is no particular limitation on its type.

[0139] Meanwhile, the above detailed description should not be construed as limiting in any respect and should be considered illustrative. The scope of the present invention should be determined by a reasonable interpretation of the appended claims, and all modifications within the equivalent scope of the present invention are intended to be included within the scope of the present invention.

Claims

1. A method for creating a 3D object in a 3D object creation system, A step of selecting one of the 3D asset creation mode and the 3D character creation mode from a user terminal through an interface provided in the above system; A step of providing a generation page corresponding to the generation mode selected in the user terminal; A step of selecting one of a plurality of detailed generation models dependent on the selected generation mode from the user terminal on the generation page; A step of receiving an input image from the user terminal; A method for generating a 3D object, characterized by including a step of generating a 3D object according to the selected detailed generation model using a target object included in the input image.

2. In paragraph 1, A method for generating a 3D object, characterized in that the method further comprises the step of generating a file name of the input image using a point in time specified as the input image, and storing the input image in a database of the system with the generated file name.

3. In paragraph 2, The steps for creating the above three-dimensional object are: A step of generating a file name of the above three-dimensional object is included, The file name of the above 3D object is, A method for generating a 3D object, characterized in that it includes unique information of a specific detailed generation model that generated the 3D object among the plurality of detailed generation models and at least a portion of the file name of the input image.

4. In paragraph 3, The above 3D character creation mode is, Including a first detailed generation model and a second detailed generation model, A three-dimensional object generation method, wherein the first detailed generation model and the second detailed generation model are different in at least one of the generation speed and the generation quality for the three-dimensional object.

5. In paragraph 3, The above 3D character creation mode is, For the target object, at least one of a first detailed generation model for generating 2D stylization, a second detailed generation model for generating a re-viewpoint view, and a third detailed generation model for performing 3D model deformation is included. A method for generating a 3D object, characterized in that the file names of the 3D object are different depending on which of the first detailed generation model, the second detailed generation model, and the third detailed generation model the 3D object was generated using in the 3D character generation mode.

6. In paragraph 3, When the 3D asset creation mode is selected from the user terminal, a creation page of the 3D asset creation mode is provided to the user terminal, In one area of ​​the creation page of the above 3D asset creation mode, comprising an object selection area including at least one 3D character generated in the above 3D character generation mode; A method for generating a 3D object, characterized by generating an extended 3D object using the 3D object and a specific 3D character selected from the object selection area.

7. In paragraph 6, The file name of the above 3D object is Further including user account information logged into the above user terminal, A method for generating a 3D object, characterized in that at least one 3D character including the user account information as a file name among a plurality of 3D characters stored in the system is loaded into the object selection area.

8. In paragraph 6, The above extended 3D object includes the above 3D object and the above specific 3D character, In the above extended 3D object, the relative positional relationship between the 3D object and the specific 3D character is A method for creating a 3D object, characterized in that it is specified based on a user input received from the user terminal for the creation page of the above 3D asset creation mode. In a 9.3D object creation system, A communication unit for selecting one of the 3D character creation modes and the 3D character creation mode from a user terminal through an interface provided in the above system; and Including a control unit that provides a generation page corresponding to the generation mode selected in the user terminal, The above control unit, From the user terminal, one of a plurality of detailed generation models dependent on the selected generation mode is selected on the generation page, Receiving an input image from the user terminal, A three-dimensional object generation system characterized by generating a three-dimensional object according to the selected detailed generation model using a target object included in the input image.

10. A program that is executed by one or more processes in an electronic device and stored in a computer-readable recording medium, The above program is, A step of selecting one of the 3D character creation modes and the 3D character creation mode from a user terminal through an interface provided in the system; A step of providing a generation page corresponding to the generation mode selected in the user terminal; A step of selecting one of a plurality of detailed generation models dependent on the selected generation mode from the user terminal on the generation page; A step of receiving an input image from the user terminal; A program stored on a computer-readable recording medium, characterized in that it includes commands for performing a step of generating a three-dimensional object according to the selected detailed generation model using a target object included in the input image.

Images