首页> 外国专利> Method and System for restoring objects and background and creating your own character for reality-based Social Network Services

Method and System for restoring objects and background and creating your own character for reality-based Social Network Services

机译:用于恢复对象和背景的方法和系统以及为基于现实的社交网络服务创建自己的字符

摘要

A method and system for creating characters and restoring objects and backgrounds for creating and sharing AR (Augmented Reality) and VR (Virtual Reality) content capable of multilateral activities are provided. A method for generating a character and restoring an object and a background according to an embodiment of the present invention includes: generating, by a character creation and restoration of an object and a background, a basic template model based on external appearance information of the object; generating, by the character creation and object and background restoration system, virtual reality and augmented reality content by applying the posture information of the object to the basic template model; and sharing, by the character creation and object and background restoration system, the generated virtual reality and augmented reality contents to a social network service (SNS). Thereby, the application of super-resolution sensor data and learning-based motion to the creation of human and animal characters, various plants or background objects is considered, and the time and cost required for creating human and animal characters, various plants or background objects can be reduced. , it can express a space very similar to the real world, and data augmentation techniques and deep learning technology can collect new 3D appearances and postures for various 2D single or multiple image inputs, making it easier to compose AR and VR content. can
机译:提供了一种用于创建字符和恢复对象的方法和系统以及用于创建和共享AR(增强现实)和能够提供多边活动的VR(虚拟现实)内容的方法和系统。一种用于生成字符和恢复对象的方法和根据本发明的实施例的背景包括:通过对象的字符创建和恢复对象和背景,基于对象的外观信息的基本模板模型来生成基本模板模型;通过将对象的姿势信息应用于基本模板模型,通过字符创建和对象和对象和背景恢复系统生成虚拟现实和增强现实内容;并通过字符创建和对象和背景恢复系统共享,生成的虚拟现实和增强现实内容到社交网络服务(SNS)。因此,考虑了超分辨率传感器数据和基于学习的运动来创建人和动物人物的创建,以及创建人类和动物人物,各种植物或背景对象所需的时间和成本可以减少。 ,它可以表达与现实世界非常相似的空间,数据增强技术和深度学习技术可以为各种2D单个或多个图像输入收集新的3D外观和姿势,从而更容易构成AR和VR内容。能够

著录项

  • 公开/公告号KR20210073428A

    专利类型

  • 公开/公告日2021-06-18

    原文格式PDF

  • 申请/专利权人 한국전자기술연구원;

    申请/专利号KR1020200028272

  • 发明设计人 김성흠;윤주홍;류한나;이민호;

    申请日2020-03-06

  • 分类号G06Q50/10;G06N3/08;G06Q50;G06T13/40;G06T19;

  • 国家 KR

  • 入库时间 2022-08-24 19:49:27

相似文献

  • 专利
  • 外文文献
  • 中文文献
获取专利

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号