【作者】 闫剑锋； 李战怀； 陈世亮；

【Author】 YAN Jian-Feng~1,LI Zhan-Huai~1,and CHEN Shi-Liang~(1,2) 1(School of Computer Science and Technology,Northwestern Polytechnical University,Xi’an 710072) 2(College of Missile,Air Force Engineering University,Xianyang 713800)

【机构】 西北工业大学计算机科学与技术学院； 空军工程大学导弹学院；

【摘要】 时态特征是视频、音频类多媒体数据具有的重要特征之一,相关研究工作主要集中在基本时态关系之上,难于回答具有时态语义抽象特征的多媒体数据查询,研究了概化和聚集这两类重要时态语义抽象的含意和描述算法,并通过例子解释了基于以上描述的多媒体时态抽象.更多还原

【Abstract】 The temporal feature is one of the most important features in multimedia data,such as video and audio.Recently,researches on content descriptions and queries of temporal features are based on the thirteen kinds of temporal relationships defined by Allen,for which it is difficult to answer the queries of multimedia data that are built on temporal semantic abstraction.Based on the definition of multimedia object and multimedia data unit(MDU),this paper focuses on the descriptive algorithm of generalization and aggregation,which are two kinds of important temporal semantic abstraction.Furthermore,the multimedia data model is introduced by using a supporting video example.更多还原