计算机工程与应用 ›› 2023, Vol. 59 ›› Issue (15): 290-299.DOI: 10.3778/j.issn.1002-8331.2205-0019

• 工程与应用 • 上一篇    下一篇

增强现实环境快速重建系统研究

彭睿男,万韬阮,巩林明   

  1. 1.西安工程大学 计算机科学学院,西安 710600
    2.新型网络智能信息服务国家地方联合工程研究中心,西安 710600
    3.陕西省服装设计智能化重点实验室,西安 710600
    4.布拉德福德大学 工程与信息学院,西约克郡 布拉德福德市 BD7 1DP
  • 出版日期:2023-08-01 发布日期:2023-08-01

Research on Rapid Reconstruction System for Augmented Reality Environment

PENG Ruinan, WAN Taoruan, GONG Linming   

  1. 1.School of Computer Science, Xi’an Polytechnic University, Xi’an 710600, China
    2.State and Local Joint Engineering Research Center for Advanced Networking & Intelligent Information Services, Xi’an 710600, China
    3.Shaanxi Key Laboratory of Clothing Intelligence, Xi’an 710600, China
    4.Faculty of Engineering and Informatics, University of Bradford, Bradford BD7 1DP, UK
  • Online:2023-08-01 Published:2023-08-01

摘要: 针对现有增强现实技术在进行环境重建时存在环境重建时间长、设备门槛高、户外环境重建定位困难等问题,设计并实现了AR环境快速重建系统。以谷歌ARCore作为开发基础,调用环境信息估算工具获取三维模型空间坐标和环境深度信息;设计一种新的多重目标跟踪匹配算法,以解析目标特征,并与三维模型生成对应关系;加入北斗全球卫星信号(global navigation satellite system,GNSS)接收机对原有的定位模块进行改进,采用参数转换法对高精度坐标进行本地化处理;使用分割位姿匹配法校正环境中的三维模型重建位姿。实验结果证明,该改进系统适用范围广,拥有脱离摄像头后AR环境不丢失位置等特点,增强了AR环境重建中用户的方位感知能力,降低了AR环境中不同三维模型和目标特征的匹配处理难度。

关键词: 增强现实, 谷歌ARCore, 多重目标跟踪匹配, 北斗全球卫星, 参数转换法, 分割位姿匹配

Abstract: In response to the problems of long environment reconstruction time, high equipment threshold and difficulty in positioning outdoor environment reconstruction when performing environment reconstruction with existing augmented reality technology, the AR environment rapid reconstruction system is designed and implemented. Using Google ARCore as the basis for development and calling on environmental information estimation tools to obtain 3D model spatial coordinates and environmental depth information. A new multiple target tracking and matching algorithm is designed to parse target features and generate correspondence with the 3D model. The BeiDou global navigation satellite system(GNSS) receiver is added to improve the original positioning module, and the parameter conversion method is used to localize the high-precision coordinates. Finally, a split pose matching method is used to correct the reconstructed pose of the 3D model in the environment. Experimental results show that the improved system has a wide range of applications, and has the characteristics of not losing the position of the AR environment after leaving the camera, it enhances the user’s orientation perception ability in the AR environment reconstruction, and reduces the difficulty of matching different 3D models and target features in the AR environment.

Key words: augmented reality, Google ARCore, multiple target tracking and matching, BeiDou global navigation satellite, parametric conversion method, split pose matching