Pet reconstruction system for setting virtual world to base configuration and its execution method

Abstract

To provide a pet reconstruction system for setting a virtual world to a base configuration, and to provide its execution method.SOLUTION: A pet reconstruction system for setting a virtual world to a base configuration and its execution method are constituted mainly of a server or one or multiple terminal devices. In the server, information is connected to each of a model database, a custom database, a data processing module, a virtual animal generation module, an environment generation module, and a real-time analysis module. By analyzing characteristics of a specific animal and in comparison with a model stored in the model database, a virtual identify having the specific animal is generated. Finally, at least one instruction is received by the real-time analysis module, and a model of a virtual identity is executed for the virtual identity on the basis of the instruction. By realizing a real-time adjustment of action of a virtual pet, connection of feeling between an owner and a pet is reconstructed in the virtual environment.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a pet reconstruction system based on a virtual world and a method for implementing the system. [Background technology]

[0002] In the past, when a pet passed away, owners could only reminisce about their pet's memories through videos and photos, and no interactive experience was possible. Traditional methods of reminiscing were limited to static visual and auditory elements, lacking dynamic elements and failing to provide an interactive experience that evokes emotional resonance. In this situation, owners were unable to realistically relive the moments they shared with their pet. Furthermore, the lack of an interactive reminiscing method meant that while owners missed their pets, they also felt a sense of distance and emptiness.

[0003] Among conventional patent documents, for example, Patent Document 1 below discloses a method, device, electronic device, and storage medium for generating a virtual pet. The method includes the steps of acquiring a plurality of pet videos, the pet videos being pet videos taken by a user from a plurality of shooting angles, the plurality of shooting angles corresponding to the plurality of pet videos, extracting characteristics from the plurality of pet videos to obtain characteristics of each part of the pet's body, determining an appearance image of the virtual pet based on the characteristics of each part of the pet's body, determining voice parameters of the virtual pet based on the user's characteristics, the voice parameters being voice parameters used when the virtual pet interacts with the user, and generating the virtual pet based on the appearance image and the voice parameters. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Chinese Patent Application Publication No. 114870404A Summary of the Invention [Problem to be solved by the invention]

[0005] However, conventional virtual pets still lack precise control over specific actions and behaviors in different situations, such as the lack of each pet's unique personality, habits, and behaviors. Conventional virtual pet systems are limited in the actions of virtual pets and are unable to realistically simulate the rich range of actions and reactions that a pet may exhibit in different situations. Therefore, how to enable a virtual pet to realistically interact with a user, the environment, or other virtual objects in different situations, thereby providing a way to recreate contact with a pet in a virtual environment, and how to enable owners to dynamically interact more directly with their deceased pets and reminisce about them in the virtual world, remains a problem that needs to be solved.

[0006] The present invention has been made in view of the above-mentioned problems in the prior art. In order to solve the above-mentioned problems, the main object of the present invention is to provide a pet reconstruction system and a method for reconstructing a pet in a virtual world, which uses a virtual world as a base, recreates a lifelike virtual pet in front of the owner's eyes, and generates interactions, thereby satisfying users' needs for customizing their virtual pet and realizing realistic interactions. The system reconstructs an emotional connection between an owner and a pet in a virtual environment, and the method for reconstructing a pet in a virtual world. [Means for solving the problem]

[0007] To achieve the above objectives, the present invention provides a virtual world based on a situation, and the system mainly comprises a server and one or more terminal devices. The server is connected (communicatively) to a model database, a custom database, a data processing module, a virtual animal generation module, an environment generation module, and a real-time analysis module. The model database is used to store multiple 3D appearance models, multiple behavior models, and multiple voice models of different types of animals. The custom database is used to store specific appearance characteristics, specific behavior characteristics, and specific voice characteristics of a specific animal. The data processing module is used to analyze the specific appearance characteristics, specific behavior characteristics, and specific voice characteristics of the specific animal and compare them with the multiple 3D appearance models, multiple behavior models, and multiple voice models in the model database, and then generate a specific 3D appearance model, specific behavior model, and specific voice model of the specific animal, which is then stored in the custom database. The virtual animal generation module generates a virtual identity of the specific animal based on the specific 3D appearance model, specific behavior model, and specific voice model of the specific animal, completing the reconstruction. The environment generation module is used to generate a virtual environment and transmit at least one environment command to the real-time analysis module, which is used to receive at least one user command from the multiple terminal devices and / or at least one environment command from the environment generation module, and executes any one or a combination of a specific three-dimensional appearance model, a specific behavior model, and a specific voice model for the virtual identity based on the at least one user command and / or the at least one environment command.In addition, one or more terminal devices are connected (communicatively connected) to a server and download a virtual identity of a specific animal, including a specific three-dimensional appearance model, a specific behavior model, and a specific voice model, thereby realizing real-time adjustment of the virtual pet's behavior, expressing more realistic reactions in different environments, satisfying users' needs for customizing virtual pets to realize realistic interactions, and reconstructing the emotional bond between owner and pet in the virtual environment.

[0008] At least the following points will become clear from the description and drawings to be described later. [Brief explanation of the drawings]

[0009] [Figure 1] 1 is a block diagram showing a schematic configuration of a pet reconstruction system based on a virtual world according to an embodiment of the present invention; [Figure 2] 1 is a schematic diagram showing the configuration of a pet reconstruction system that uses a virtual world as a base configuration in one embodiment of the present invention. FIG. [Figure 3] FIG. 1 is a schematic diagram (1) showing an embodiment of a method for implementing a pet reconstruction system based on a virtual world according to the present invention. [Figure 4] FIG. 2 is a schematic diagram (2) showing one embodiment of a method for implementing a pet reconstruction system based on a virtual world according to the present invention. [Figure 5] FIG. 1 is a schematic diagram (1) showing the implementation status of the virtual world-based pet reconstruction system according to the present invention. [Figure 6] FIG. 3 is a schematic diagram (3) showing an embodiment of a method for implementing a pet reconstruction system based on a virtual world according to the present invention. [Figure 7] FIG. 2 is a schematic diagram (II) showing the implementation of the method for the pet reconstruction system based on the virtual world according to the present invention. [Figure 8] FIG. 4 is a schematic diagram (4) showing an embodiment of a method for implementing a pet reconstruction system based on a virtual world according to the present invention. [Figure 9]FIG. 3 is a schematic diagram (III) showing the implementation of the method for constructing a pet reconstructing system based on a virtual world according to the present invention. [Figure 10] FIG. 5 is a schematic diagram (5) showing an embodiment of a method for implementing a pet reconstruction system based on a virtual world according to the present invention. [Figure 11] FIG. 4 is a schematic diagram (4) showing the implementation of the method for constructing a pet reconstructing system based on a virtual world according to the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0010] The present invention will be described below through embodiments of the invention, but the following embodiments do not limit the scope of the invention according to the claims. Furthermore, not all of the combinations of features described in the embodiments are necessarily essential to the solution of the invention.

[0011] The pet reconstruction system 1 according to the present invention is based on a virtual world as a situation, and is comprised of a server 11 and one or more terminal devices 13 (see FIG. 1). The server 11 is connected (communicatively connected) to a model database 111, a custom database 112, a data processing module 113, a virtual animal generation module 114, an environment generation module 115, and a real-time analysis module 116. The model database 111 is used to store multiple three-dimensional appearance models 1111, multiple behavior models 1112, and multiple voice models 1113 of different types of animals. The custom database 112 is used to store specific appearance characteristics 1121, specific behavior characteristics 1122, and specific voice characteristics 1123 of specific animals. The data processing module 113 is used to analyze the specific appearance characteristics 1121, the specific behavioral characteristics 1122, and the specific sound characteristics 1123 of the specific animal, compare them with the multiple three-dimensional appearance models 1111, the multiple behavioral models 1112, and the multiple sound models 1113 in the model database 111, and then generate a specific three-dimensional appearance model 1121′, a specific behavioral model 1122′, and a specific sound model 1123′ of the specific animal, and store them in the custom database 112. The virtual animal generation module 114 generates a virtual identity 1141 of the specific animal based on the specific three-dimensional appearance model 1121′, the specific behavioral model 1122′, and the specific sound model 1123′ of the specific animal, completing the reconstruction. The environment generation module 115 is used to generate a virtual environment 1151 and send at least one environment command to the real-time analysis module 116. The real-time analysis module 116 is used to receive at least one user command from a plurality of terminal devices 13 and / or at least one environment command from the environment generation module 115, and executes any one or combination of a specific three-dimensional appearance model 1121′, a specific behavior model 1122′, and a specific voice model 1123′ for the virtual identity 1141 based on the at least one user command and / or the at least one environment command.Alternatively, the real-time analysis module 116 causes the virtual identity 1141 to send and / or receive at least one virtual animal command from at least one other virtual identity 1141′ in the virtual environment 1151, and causes the virtual identity 1141 to execute any one or a combination of a specific three-dimensional appearance model 1121′, a specific behavior model 1122′, and a specific voice model 1123′ based on the at least one virtual animal command. Note that one or more terminal devices 13 are informationally connected (communicatively connected) to the server 11 and download the virtual identity 1141 including the specific three-dimensional appearance model 1121′, the specific behavior model 1122′, and the specific voice model 1123′ of the specific animal.

[0012] The three-dimensional appearance model 1111 stored in the model database 111 and the specific three-dimensional appearance model 1121′ stored in the custom database 112 each include at least one or a combination of parameters such as breed, sex, age, skin color, coat color, pattern, body shape, and external structure, but the present invention is not limited thereto. The behavior model 1112 and the specific behavior model 1122′ are at least one or a combination of parameters of habits or actions such as sleeping, eating, communication, entertainment, learning, attack, defense, etc. The voice model 1113 and the specific voice model 1123′ each include at least one or a combination of parameters such as tone, frequency, volume, rhythm, continuity, and voice type, but the present invention is not limited thereto.

[0013] The terminal device 13 includes a camera 131 and a microphone 132. The camera 131 captures real-time video data of the real world and transmits it to the server 11. The microphone 132 captures real-time audio data of the real world and transmits it to the server 11. The camera 131 also includes a motion detection module 1311, which captures motion data of at least one object in the real-time video data and transmits it to the server 11.

[0014] In the present invention, the real-time analysis module 116 further adjusts the various parameters of the specific three-dimensional appearance model 1121′, the specific behavior model 1122′, and the specific voice model 1123′ of the virtual identity 1141 based on the environmental command of the virtual environment 1151. In addition, the real-time analysis module 116 adjusts the various parameters of the specific three-dimensional appearance model 1121′, the specific behavior model 1122′, and the specific voice model 1123′ of the virtual identity 1141 based on the real-time video data transmitted to the server 11 by the camera 131, the real-time audio data transmitted to the server 11 by the microphone 132, and / or the movement data transmitted to the server 11 by the motion detection module 1311.

[0015] In this case, as shown in FIG. 2 , a server 11 is connected (communicatively connected) to one or more terminal devices 13 via the Internet 12, and users of the terminal devices 13 download necessary data from the server 11 or transmit necessary data to the server 11 via the Internet 12. The Internet 12 may be a broadband network, a 4G network, a 5G network, a Wi-Fi network, or a combination thereof. The terminal device 13 may be a mobile phone, a tablet device, a personal computer, or any other device that can connect to the server 11 via a network. The terminal device 13 is used in combination with a virtual reality display device 14, which may be a goggle-type or head-mounted virtual reality display device and displays all virtual object information downloaded to the terminal device 13.

[0016] The implementation method according to the present invention mainly includes the following steps (see FIG. 3). (1) “Data Upload S1” step: Upload specific appearance characteristics 1121, specific behavioral characteristics 1122, and specific sound characteristics 1123 of a specific animal to the custom database 112 of the server 11. (2) “Model Construction S2” step: The data processing module 113 analyzes the specific appearance characteristics 1121, the specific behavioral characteristics 1122, and the specific sound characteristics 1123 of the specific animal, and generates a specific three-dimensional appearance model 1121′, a specific behavioral model 1122′, and a specific sound model 1123′ of the specific animal, and stores them in the custom database 112. (3) “Virtual Animal Generation S3” step: The virtual animal generation module 114 generates a virtual identity 1141 of the specific animal based on the specific three-dimensional appearance model 1121′, the specific behavior model 1122′, and the specific voice model 1123′ of the specific animal, thereby completing the reconstruction. (4) "Environment Generation S4" Step: The environment generation module 115 generates a virtual environment 1151. (5) “Interaction S5” step: The real-time analysis module 116 receives at least one user command from the terminal device 13 and / or at least one environment command from the environment generation module 115, and executes any one or a combination of a specific three-dimensional appearance model 1121′, a specific behavior model 1122′, and a specific voice model 1123′ for the virtual identity 1141 based on the at least one user command and / or the at least one environment command. Alternatively, the real-time analysis module 116 causes the virtual identity 1141 to send and / or receive at least one virtual animal command from at least one other virtual identity 1141' in the virtual environment 1151, and causes the virtual identity 1141 to execute any one or combination of a specific three-dimensional appearance model 1121', a specific behavior model 1122', and a specific voice model 1123' based on the at least one virtual animal command, and adjusts various parameters of the specific three-dimensional appearance model 1121', the specific behavior model 1122', and the specific voice model 1123' of the virtual identity 1141, and performs corresponding actions.

[0017] The "upload" in the data upload S1 step may, for example, provide a software application, allow a user to run the software application on a terminal device 13, input specific appearance characteristics 1121, specific behavioral characteristics 1122, and specific audio characteristics 1123 of the specific animal, and upload the data to the custom database 112 on the server 11 via the Internet 12. Alternatively, a website interface may be provided, allowing a user to input the characteristics via a browser, and upload the data to the custom database 112 on the server 11 via the Internet 12. Note that the means for uploading via the software application or website interface may further include an API (application programming interface), allowing a developer to write an API request using a programming language (e.g., Python, Java), and uploading the characteristics to the custom database 112 on the server 11 in a specific format (e.g., JSON, XML, etc.) via the API request. Alternatively, for example, data including the characteristics may be stored on a conventional external hard disk (e.g., a solid-state drive (SSD) or hard disk drive (HDD)), and the characteristics may be directly uploaded by directly connecting the external hard disk to a physical device where the server 11 is located.

[0018] In addition, in the data upload S1 step, an animal wearable device is further provided that is equipped with an image sensor (camera, etc.) for capturing specific appearance characteristics 1121 of the specific animal, a behavioral sensor (accelerometer, gyro, etc.) for capturing specific behavioral characteristics 1122 of the specific animal, and an audio sensor (microphone, etc.) for capturing specific audio characteristics 1123 of the specific animal, and the data is uploaded to the custom database 112 of the server 11 via the Internet 12, but the present invention is not limited to this.

[0019] In this case, as shown in Fig. 4, the model construction step S2 further includes a "model optimization step S21", in which the data processing module 113 first compares the specific appearance characteristics 1121, the specific behavioral characteristics 1122, and the specific voice characteristics 1123 of the specific animal with the multiple 3D appearance models 1111, the multiple behavioral models 1112, and the multiple voice models 1113 of different kinds of animals stored in the model database 111, and then generates a specific 3D appearance model 1121', a specific behavioral model 1122', and a specific voice model 1123' of the specific animal, thereby further improving the completeness and accuracy of the model.

[0020] In addition, in the model optimization step S21, deep learning such as generative adversarial networks (GANs) is further used to achieve model optimization, which includes the following steps: (1) "Data Preprocessing G1" step: The specific appearance characteristics 1121, specific behavioral characteristics 1122, and specific voice characteristics 1123 of a specific animal collected in the data upload S1 step, and / or other supplementary data (e.g., a text description of a pet's behavior dictated by the owner), as well as the multiple 3D appearance models 1111, multiple behavioral models 1112, and multiple voice models 1113 stored in the model database 111, are subjected to data cleansing, standardization, and preprocessing to ensure data consistency and to convert the data into real animal data. (2) "Create Generator G2" step: A generator model is created to receive a random noise vector as input and output a prediction result of the specific animal type. For example, if the generator generates characteristics of a canine based on the random noise vector, slight unnatural prediction results may occur in appearances such as coat texture, color variations, and the shape of certain parts. (3) "Create discriminator G3" step: A discriminator model is created to represent the input, which is real animal data or inference results, and to further distinguish between real animal data and inference results generated by the generator. (4) "G4: Training the GAN model" step. The inferences generated by the generator and real animal data are used to train the classifier, which learns to distinguish between real animal data and the inferences. At the same time, the generator is trained to trick the classifier, making the inferences it generates more realistic. (5) “GAN model optimization G5” step: Repeat the G4 step to train the GAN model multiple times, adjusting the weights of the generator and discriminator as needed to optimize the effectiveness of the GAN model.

[0021] 5, if the characteristic data of the specific animal provided is insufficient, i.e., if only partial animal characteristic data is uploaded, the data processing module 113 of the server 11 searches for available similar models in the model database 111 through a comparison process in the model optimization step S21 to fill in the missing characteristics of the specific animal. Specifically, for example, if the provided characteristic data is missing the complete appearance, body, or limbs of the specific animal, or if the voice or unique habits of the specific animal are missing, the data processing module 113 can infer the missing characteristics through a comparison process and use data of individuals of similar animal species in the model database 111 to further improve the completeness and accuracy of the model, thereby constructing a realistic virtual identity 1141 even in a situation where the characteristic data of the specific animal is insufficient, and providing the user with an overall more realistic experience.

[0022] In this case, as shown in Fig. 6, the interaction step S5 further includes a "real-time analysis step S51". The real-time analysis module 116 adjusts the various parameters of the specific three-dimensional appearance model 1121', the specific behavior model 1122', and the specific voice model 1123' of the virtual identity 1141 based on the real-time video data transmitted to the server 11 by the camera 131, the real-time audio data transmitted to the server 11 by the microphone 132, and / or the movement data transmitted to the server 11 by the motion detection module 1311, thereby allowing the virtual identity 1141 to interact more closely with the user and providing the user with a more dynamic, personalized, and realistic experience.

[0023] In addition, the real-time analysis step S51 further includes the following steps. (1) “Detecting user behavior U1” step: The camera 131, the microphone 132, and / or the motion detection module 1311 detect the user’s behavior, such as one or a combination of movements, hand movements, and / or voice, and transmit the detected behavior to the server 11, respectively. (2) “User Data Collection U2” step: The real-time analysis module 116 collects real-time user video data, real-time user audio data, and / or user motion data corresponding to the user's actions in the virtual environment 1151. (3) "User Data Analysis U3" step: The real-time analysis module 116 performs sentiment analysis on the user data to understand the user's intentions and commands. For example, through GAN deep learning, a large amount of data (e.g., facial expressions, speech tones, or hand movements) containing positive, negative, or neutral sentiment tags is collected in advance as training samples for the deep learning model, allowing the real-time analysis module 116 to predict the sentiment type of a specific action or voice. (4) “Generate user commands U4” step: Based on the emotion analysis result, the real-time analysis module 116 generates corresponding user commands for the virtual identity 1141. (5) "U51: Adjusting Behavior" step: After the virtual identity 1141 receives the user's command based on the emotion analysis result, the real-time analysis module 116 adjusts the parameters of the specific three-dimensional appearance model 1121', the specific behavior model 1122', and the specific voice model 1123', so that the virtual identity 1141 performs a corresponding behavior (e.g., sleeping, eating, communicating, etc.). Alternatively, in the "U52: Adjusting Emotion Expression" step, the real-time analysis module 116 adjusts the parameters of the specific three-dimensional appearance model 1121', the specific behavior model 1122', and the specific voice model 1123', so that the virtual identity 1141 expresses a corresponding emotion (e.g., excitement, fear, happiness, etc.).

[0024] 6 and 7, when a user performs a handshake motion, the real-time analysis module 116 captures the user's visual actions through the camera 131, detects the hand movements and positions, and determines that the user is performing a handshake motion. Alternatively, the real-time analysis module 116 receives an audio signal through the microphone 132 and captures other possible command voices (e.g., verbal instructions to shake hands). Then, the virtual identity 1141 performs the corresponding handshake motion through steps U1 to U52.

[0025] In this case, as shown in Figure 8, in the real-time analysis S51 step, various parameters of the specific three-dimensional appearance model 1121', the specific behavior model 1122', and the specific voice model 1123' of the virtual identity 1141 are further adjusted according to the environmental command of the virtual environment 1151. The real-time analysis S51 step includes the following steps: (1) "Environmental Data Collection E1" step: The real-time analysis module 116 collects environmental change data of the virtual environment 1151, including the situation, day / night, weather changes, changes in light, movement of objects, and the like. (2) "Environmental Data Analysis E2" step: The real-time analysis module 116 performs an environmental analysis on the environmental change data to determine trigger conditions. For example, using GAN deep learning, a large amount of data including environmental parameters (e.g., the volume of thunder and rain, the brightness of sunlight and lightning, the amplitude of ground vibrations, etc.) is collected in advance as training samples for the deep learning model, allowing the real-time analysis module 116 to determine whether the virtual identity 1141 will trigger a specific behavior or emotion in a specific environment. (3) "Generate environmental commands E3" step: The real-time analysis module 116 generates corresponding environmental commands for the virtual identity 1141 based on the analysis result of the environment. (4) "Adjusting Behavior E41" step: After the virtual identity 1141 receives the environmental command based on the analysis result of the environment, the real-time analysis module 116 adjusts the various parameters of the specific three-dimensional appearance model 1121', the specific behavior model 1122', and the specific voice model 1123', so that the virtual identity 1141 performs a corresponding behavior (e.g., sleeping, eating, communicating, etc.). Alternatively, in the "Adjusting Emotion Expression E42" step, the real-time analysis module 116 adjusts the various parameters of the specific three-dimensional appearance model 1121', the specific behavior model 1122', and the specific voice model 1123', so that the virtual identity 1141 expresses a corresponding emotion (e.g., excitement, fear, happiness, etc.).

[0026] Next, referring to Figures 8 and 9, when the environment generation module 115 introduces the weather effects of virtual dark clouds, lightning, and thunder into the virtual environment 1151, the real-time analysis module 116 triggers the virtual identity 1141 to feel fear through steps E1 to E42, and the virtual identity 1141 correspondingly covers both eyes with its front legs.

[0027] 10 , in the real-time analysis S51 step, the various parameters of the specific three-dimensional appearance model 1121′, the specific behavior model 1122′, and the specific voice model 1123′ of the virtual identity 1141 are further adjusted based on the behavior and emotion of at least one other virtual identity 1141′ in the virtual environment 1151. The real-time analysis S51 step includes the following steps: (1) "Virtual Animal Data Collection O1" step: The real-time analysis module 116 collects behavioral and emotional data, including voice, movement, position, etc., of other virtual identities 1141' in the virtual environment 1151. (2) "Virtual Animal Data Analysis O2" step: The real-time analysis module 116 analyzes the behavioral and emotional data of the other virtual identities 1141' to determine trigger conditions. For example, using GAN deep learning, a large amount of data including animal interactions (e.g., sniffing each other, playing, cooperative behavior, etc.) can be collected in advance as training samples for the deep learning model, and the real-time analysis module 116 can determine that certain behaviors and emotions trigger the virtual identity 1141 to perform certain behaviors and emotions. (3) “Generate a command for the virtual animal O3” step: Based on the analysis result of the virtual animal, the real-time analysis module 116 generates a corresponding virtual animal command for the virtual identity 1141. (4) "Adjusting behavior O41" step: After the virtual identity 1141 receives the virtual animal command based on the analysis result of the virtual animal, the real-time analysis module 116 adjusts the various parameters of the specific three-dimensional appearance model 1121', the specific behavior model 1122', and the specific voice model 1123', so that the virtual identity 1141 performs the corresponding behavior (playing, eating together, communicating, etc.). Alternatively, in the "adjusting emotion expression O42" step, the real-time analysis module 116 adjusts the various parameters of the specific three-dimensional appearance model 1121', the specific behavior model 1122', and the specific voice model 1123', so that the virtual identity 1141 expresses the corresponding emotion (excitement, fear, happiness, etc.).

[0028] Next, referring to Figures 10 and 11 together, when the real-time analysis module 116 sends a specific signal from the virtual identity 1141 to other adjacent virtual identities 1141', the real-time analysis module 116 triggers the virtual identity 1141 to perform a sociable behavior and executes a corresponding reaction through steps O1 to O42.

[0029] As described above, the implementation of the present invention reliably achieves the objective of providing a pet reconstruction system that uses a virtual world as a base, recreates a lifelike virtual pet in front of the owner's eyes, generates interactions, and meets the user's demand for realistic interactions with customized virtual pets, thereby reconstructing the emotional connection between the owner and the pet in a virtual environment, and a method for reconstructing a pet in a virtual world.

[0030] Although the embodiment of the present invention has been described above in detail with reference to the drawings, the specific configuration is not limited to this embodiment, and design changes and the like are also included within the scope that does not deviate from the gist of the present invention. [Explanation of symbols]

[0031] 1 Pet Reconstruction System 11 Server 12. Internet 111 Model Database 13 Terminal Equipment 112 Custom Database 131 Camera 113 Data Processing Module 132 Mike 114 Virtual Animal Generation Module 1311 Motion detection module 115 Environment Generation Module 14 Virtual reality display devices 116 Real-time Analysis Module 1111 3D Appearance Model 1112 Behavioral Model 1113 Voice Model 1121 Specific Appearance Characteristics 1122 Specific Behavioral Characteristics 1123 Specific Voice Characteristics 1121' Specific 3D appearance model 1122' Specific Behavioral Model 1123′ Specific voice model 1141 Virtual Identity 1141' Other Virtual Identities 1151 Virtual Environment S1 Data Upload G1 data preprocessing S2 Model Building Creating a G2 generator S3 Virtual Animal Generation Create a G3 classifier S4 environment generation G4 Training GAN Model S5 interaction Optimizing the G5 GAN model S21 model optimization S51 Real-time Analysis U1 Detect user behavior E1 Environmental data collection U2 User Data Collection E2 Environmental Data Analysis U3 User Data Analysis E3 Generate environment command U4 Generate user instructions E41 Coordinating behavior U51 Adjusting behavior E42 Adjusting Emotional Expressions U52 Adjusting Emotion Expressions O1 Virtual Animal Data Collection O2 Virtual Animal Data Analysis O3 Generate instructions for virtual animals O41 Adjusting behavior O42 Adjust emotional expression

Claims

1. A pet reconstruction system in which a virtual world is configured as a base, and the virtual world is configured as a base mainly composed of a server and one or more terminal devices, wherein the server: a model database communicatively connected to the server, the model database being used to store a plurality of 3D appearance models, a plurality of behavior models, and a plurality of sound models of different kinds of animals; a custom database communicatively connected to the server and used to store specific appearance characteristics, specific behavioral characteristics, and specific audio characteristics of specific animals that are uploaded; a data processing module communicatively connected to the server, the data processing module being adapted to analyze the specific appearance characteristics, the specific behavioral characteristics, and the specific sound characteristics of the specific animal, and then generate a specific three-dimensional appearance model, a specific behavioral model, and a specific sound model of the specific animal, and store the model in the custom database; a virtual animal generation module communicatively connected to the server, the virtual animal generation module being configured to generate a virtual identity of the specific animal based on the specific three-dimensional appearance model, the specific behavior model, and the specific voice model of the specific animal, and complete the reconstruction; and a real-time analysis module that is communicatively connected to the server and used to receive at least one user command from the terminal device, and that executes any one or a combination of the specific three-dimensional appearance model, the specific behavior model, and the specific voice model for the virtual identity based on the user command, thereby adjusting various parameters of the specific three-dimensional appearance model, the specific behavior model, and the specific voice model of the virtual identity and causing the virtual identity to perform corresponding behavior.

2. The pet reconstruction system for base-configuring a virtual world as described in claim 1, characterized in that the terminal device is communicatively connected to the server via the Internet and downloads the virtual identity including the specific three-dimensional appearance model, the specific behavior model, and the specific voice model of the specific animal.

3. The pet reconstruction system based on a virtual world, as described in claim 1, characterized in that the terminal device further comprises a camera for capturing real-time video data of the real world and a microphone for capturing real-time audio data of the real world, and the camera has an operation detection module for capturing movement data of at least one object in the real-time video data and transmitting the movement data to the server, respectively.

4. The pet reconstruction system for base-configuring a virtual world as described in claim 3, characterized in that the server further comprises an environment generation module used to generate a virtual environment and transmit at least one environment command to the real-time analysis module.

5. The pet reconstruction system for base-configuring a virtual world as described in claim 4, characterized in that the real-time analysis module receives the environmental command from the environment generation module and executes any one or combination of the specific three-dimensional appearance model, the specific behavior model, and the specific voice model for the virtual identity based on the environmental command.

6. The pet reconstruction system for base-configuring a virtual world as described in claim 4, characterized in that the real-time analysis module causes the virtual identity to send and / or receive at least one virtual animal command from at least one other virtual identity in the virtual environment, and executes any one or combination of the specific three-dimensional appearance model, the specific behavior model, and the specific voice model based on the virtual animal command.

7. 5. The pet reconstruction system for constructing a virtual world as a base, as described in claim 4, wherein each of the three-dimensional appearance models and the specific three-dimensional appearance model includes at least one or a combination of parameters of breed, sex, age, skin color, coat color, pattern, body shape, and external structure; each of the behavioral models and the specific behavioral model includes at least one or a combination of parameters of sleep, eating, communication, entertainment, learning, and attack / defense; and each of the voice models and the specific voice model includes at least one or a combination of parameters of tone, frequency, volume, rhythm, continuity, and type of voice.

8. The pet reconstruction system for base-configuring a virtual world as described in claim 7, characterized in that the real-time analysis module further adjusts the parameters of the specific three-dimensional appearance model, the specific behavior model, and the specific sound model based on the environmental command.

9. The pet reconstruction system based on a virtual world as described in claim 7, characterized in that the real-time analysis module further adjusts each of the parameters of the specific 3D appearance model, the specific behavior model, and the specific voice model based on the real-time video data, the real-time voice data, and / or the movement data transmitted to the server by the camera, the microphone, and / or the motion detection module, respectively.

10. a data uploading step of uploading specific appearance characteristics, specific behavioral characteristics, and specific audio characteristics of specific animals to a custom database on a server; a model construction step of analyzing the specific appearance characteristics, the specific behavior characteristics, and the specific sound characteristics of the specific animal by a data processing module of the server, generating a specific three-dimensional appearance model, a specific behavior model, and a specific sound model of the specific animal, and storing them in the custom database; a virtual animal generation step of generating a virtual identity of the specific animal based on the specific three-dimensional appearance model, the specific behavior model, and the specific voice model by a virtual animal generation module of the server; an environment generation step of generating a virtual environment by an environment generation module of the server; A method for implementing a pet reconstruction system that bases a virtual world on a virtual object, comprising: an interaction step of receiving, by a real-time analysis module of the server, at least one user command from one or more terminal devices communicatively connected to the server, and / or at least one environment command from the environment generation module, and executing, for the virtual identity, any one or a combination of the specific three-dimensional appearance model, the specific behavior model, and the specific voice model based on the user command and / or the environment command.

11. The method for implementing a pet reconstruction system based on a virtual world, as described in claim 10, characterized in that the terminal device further comprises a camera for capturing real-time video data of the real world and a microphone for capturing real-time audio data of the real world, and the camera has an operation detection module for capturing movement data of at least one object in the real-time video data and transmitting the movement data to the server respectively.

12. The method for implementing a pet reconstruction system based on virtual world construction, as described in claim 11, characterized in that the model construction step further includes a model optimization step in which the data processing module first compares the specific appearance characteristics, the specific behavioral characteristics, and the specific sound characteristics of the specific animal with multiple three-dimensional appearance models, multiple behavioral models, and multiple sound models of different types of animals stored in a model database of the server, and then generates the specific three-dimensional appearance model, the specific behavioral model, and the specific sound model of the specific animal.

13. 13. The method for implementing a pet reconstruction system based on a virtual world configuration, as described in claim 12, wherein the model optimization step uses a generative adversarial network (GAN).

14. 13. The method for implementing a pet reconstruction system using a virtual world as a base configuration, as described in claim 12, wherein each of the three-dimensional appearance models and the specific three-dimensional appearance model includes at least one or a combination of parameters of breed, sex, age, skin color, coat color, pattern, body shape, and external structure; each of the behavioral models and the specific behavioral model includes at least one or a combination of parameters of sleep, eating, communication, entertainment, learning, and attack / defense; and each of the voice models and the specific voice model includes at least one or a combination of parameters of tone, frequency, volume, rhythm, continuity, and type of voice.

15. The method for implementing a pet reconstruction system based on virtual world construction as described in claim 14, characterized in that the interaction step further includes a real-time analysis step of adjusting each of the parameters of the specific 3D appearance model, the specific behavior model, and the specific voice model by the real-time analysis module based on the real-time video data, the real-time voice data, and / or the movement data transmitted to the server by the camera, the microphone, and / or the motion detection module, respectively.

16. The real-time analysis step includes: sensing user actions by the camera, the microphone, and / or the motion sensing module, by sensing one or a combination of motion, hand movement, and / or voice, and transmitting the motion, hand movement, and / or voice to the server, respectively; a user data collection step of collecting, by the real-time analysis module, real-time user video data, real-time user voice data, and / or user motion data corresponding to one or a combination of the motion, the hand movement, and / or the voice; a user data analysis step of analyzing emotions regarding the user's real-time video data, the user's real-time voice data, and / or the user's motion data by the real-time analysis module; generating a user command, wherein the real-time analysis module generates a corresponding user command for the virtual identity based on the emotion analysis result; The method for implementing a pet reconstruction system based on virtual world configuration, as described in claim 15, further comprising the step of adjusting the parameters of the specific three-dimensional appearance model, the specific behavior model, and the specific voice model by the real-time analysis module after the virtual identity receives the user's command based on the emotion analysis result, thereby adjusting the behavior of the virtual identity and / or adjusting the emotional expression of the virtual identity.

17. The real-time analysis step includes: an environmental data collection step for collecting environmental change data in the virtual environment by the real-time analysis module; an environmental data analysis step for performing an environmental analysis on the environmental change data by the real-time analysis module; generating an environmental instruction by the real-time analysis module based on the analysis result of the environment, for generating a corresponding environmental instruction for the virtual identity; The method for implementing a pet reconstruction system based on virtual world configuration, as described in claim 15, further comprising the step of adjusting the parameters of the specific three-dimensional appearance model, the specific behavior model, and the specific voice model by the real-time analysis module after the virtual identity receives the environmental command based on the analysis result of the environment, thereby adjusting the behavior of the virtual identity and / or adjusting the emotional expression of the virtual identity.

18. The real-time analysis step includes: a virtual animal data collection step for collecting behavioral and emotional data of other virtual identities in the virtual environment by the real-time analysis module; a virtual animal data analysis step for performing, by the real-time analysis module, an analysis of a virtual animal based on the behavioral and emotional data of the other virtual identities; generating a virtual animal instruction, based on the analysis result of the virtual animal, by the real-time analysis module, for generating a corresponding virtual animal instruction for the virtual identity; The method for implementing a pet reconstruction system based on virtual world configuration, as described in claim 15, further comprising the steps of: after the virtual identity receives the virtual animal command based on the analysis result of the virtual animal, adjusting each of the parameters of the specific three-dimensional appearance model, the specific behavior model, and the specific voice model by the real-time analysis module, thereby adjusting the behavior of the virtual identity and / or adjusting the emotional expression of the virtual identity.

Images