A Novel Approach to Video Database Indexing via Semantic Classification

通过语义分类进行视频数据库索引的新方法

• 批准号: 0208539
• 负责人: Jianping Fan
• 金额: $ 17.68万
• 依托单位: University of North Carolina at Charlotte
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 2002
• 资助国家: 美国
• 起止时间: 2002-08-15 至 2005-07-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0208539&HistoricalAwards=false
• 关键词: Novel; Approach; Video; Database; Indexing

The goal of this research project is to develop a novel framework to support more effective and secure content-based video access over large-scale database. Specifically, the framework includes: (a) a hierarchical video database management and indexing structure to define the contextual and logical relationships among the related visual concepts and thus facilitate more efficient video access; (b) a probabilistic semantic video classifier to bridge the semantic gap between the low-level visual features and the high-level visual concepts; (c) a hierarchical visual summarization and presentation technique to enable fast video browsing over IP networks; (d) a multi-level video database access control scheme to serve different classes of users and maintain the video data integrity and security. A wide range of potential applications could benefit from this research, such as education and distance learning, health care and telemedicine, digital library, national security and defense, network-based media, and entertainment

本研究的目的是开发一个新的框架，以支持更有效和更安全的基于内容的视频访问大规模数据库。具体地，该框架包括：（a）分级视频数据库管理和索引结构，用于定义相关视觉概念之间的上下文和逻辑关系，从而促进更有效的视频访问;（B）概率语义视频分类器，用于桥接低级视觉特征和高级视觉概念之间的语义鸿沟;（c）采用分级视觉摘要及呈现技术，让市民可透过IP网络快速浏览视频;（d）采用多级视频数据库存取控制计划，为不同类别的用户提供服务，并维持视频数据的完整性和安全性。 这项研究可以在教育和远程教育、医疗保健和远程医疗、数字图书馆、国家安全和国防、基于网络的媒体和娱乐等领域得到广泛的应用