移动云计算——移动增强现实技术和服务  

Mobile Augmented Reality Technology and Service in Mobile Cloud Computing

作　　者：吕强[1] 黄成[1] 刘明[1] 

机构地区：[1]中兴通讯股份有限公司,深圳518057

出　　处：《中兴通讯技术》2015年第2期25-29,共5页ZTE Technology Journal

基　　金：国家科技重大专项(2012ZX03002004);国家高技术研究发展("863")计划(2013AA013904)

摘　　要：提出了中兴通讯移动增强现实解决方案。方案针对增强现实目标的识别技术采用客户机/服务器(C/S)架构,以充分利用云的海量存储和并行计算能力,进行样本图像的训练以及识别样本库的生成和训练,从而实现快速的增强现实目标的迅速识别;针对海量增强现实内容信息和训练样本数据高效存储的需求,采用分布式的非结构化存储机制以保存图像等信息,对于兴趣点信息采用结构化存储机制以保存兴趣点各种增强现实元数据信息;通过在结构化存储机制中建立特征索引等快速查询机制,以达到高效数据查询的目的。此外,为了分担增强现实业务请求负载,在靠近用户的地方部署若干用户接入云,缓存热点的增强现实信息,减轻增强现实业务处理机的负载压力。In this paper, we present ZTE ’s Mobile Augmented Reality system. To take advantage of mass storage and paral el computing power of the cloud, we choose the C/S architecture. This architecture trains sample images training and generates sample library for rapid AR target recognition. To meet the efficient storage requirements of massive AR Content information and related training sample, we save the images based on the distributed unstructured storage and save the augmented reality metadata, such as POI information, based on structured storage. We establish indexing in our structured storage to improve the efficiency of data queries. A number of access clouds are deployed near end users to cache the hot augmented reality information with the aim of reducing traffic load on our augmented reality system.

关 键 词：移动增强现实 增强现实系统 目标识别 云计算 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术] TP393.09[自动化与计算机技术—计算机科学与技术]