张兴旺,黄晓斌.国外移动视觉搜索研究述评[J].中国图书馆学报,2014,40(3):114~128
An Overview of Mobile Visual Search Research Abroad
国外移动视觉搜索研究述评
Received:September 10, 2013  Revised:December 03, 2013
DOI：
Key words:Mobile search  Mobile visual search  Mobile image search  Digital library
中文关键词:  移动搜索  移动视觉搜索  移动图像搜索  数字图书馆
基金项目:
Author NameAffiliationE-mail
Zhang Xingwang 桂林理工大学图书馆,广西 桂林,541004 zhangxwang@gmail.com 
Huang Xiaobin 中山大学资讯管理学院,广东 广州,510006  
Hits: 4129
Download times: 3787
Abstract:
Mobile visual search (MVS) as an important way to obtain information, has become the forefront of the field of information retrieval. Currently, the main research methods of MVS are simulation, comparative analysis, interdisciplinary research, field survey. The emergence of MVS will influence the modes of knowledge interaction and knowledge service, affect the search engine market share, and give rise to new industrial chains and industrial clusters. The basic architecture of MVS can be divided into standard architectures, localization architecture and hybrid architecture, involving descriptor processing techniques, visual object matching technology, visual object retrieval processes, visual object repository construction and other key technologies. The current main technical bottlenecks are:matching hardware and software resources, adaptive problems between diversity of visual query and MVS services and applications, matching the performance of MVS and user experience, interoperability problems of services, applications and heterogeneous MVS system. LIS workers should focus on issues as follows:MVS-support information retrieval mode, visual object repository construction, MVS systems and standardized visual resources, MVS application analysis and decision support, personnel training for MVS development, application and management. 3figs. 1tab. 69refs.
中文摘要:
      移动视觉搜索(MVS)作为一种重要的信息获取方式,已成为信息检索领域的前沿课题。目前学界对于MVS的研究方法主要有模拟仿真法、比较分析法、文献研究法、跨学科研究法、实地调查法等。MVS的出现将影响知识交互和知识服务模式,影响搜索引擎市场份额,并催生新型产业链及产业集群。MVS可分为标准架构、本地化架构和混合架构,涉及描述符处理技术、视觉对象对匹配技术、视觉对象检索流程、视觉对象知识库建设等关键技术。当前主要技术瓶颈有:软硬件资源的匹配问题,视觉查询多样性与MVS服务、应用的自适应问题,MVS搜索性能与用户体验效果的匹配问题,多样化移动视觉服务、应用与异构MVS系统之间的互操作问题。图情工作者应重点关注以下内容:支持MVS的信息检索模式,视觉对象知识库建设,MVS系统及视觉资源标准化,MVS应用分析及决策支持,MVS开发、应用及管理人才培养。图3。表1。参考文献69。
View Full Text   View/Add Comment  Download reader