基于视频对象的视频内容分级描述模型  

Video Content Hierarchical Description Model Based on Video Objects

作　　者：于跃龙[1] 刘志刚[1] 卢焕章[1] 

机构地区：[1]国防科技大学ATR重点实验室,长沙410073

出　　处：《微处理机》2005年第6期57-59,共3页Microprocessors

摘　　要：本文提出了一种基于视频对象的视频内容分级描述模型。视频序列首先被分成一个个的镜头,在每个镜头内对视频对象进行分割和跟踪。按照镜头、视频对象、视频对象平面和视频对象区域四级框架提取特征,对视频内容进行分级描述。本文对视频内容信息的描述可以用于视频检索、视频注释等应用。This paper brings out a video content hierarchical description model based on video objects. Video sequences are divided into shots first, in which video objects segmentation and tracking are implemented. Video content is described by characteristics extracted on four-level framework including shots, video objects, video object planes and video object regions. The description proposed in this paper can be used in various applications such as video retrieval and video annotation.

关 键 词：视频内容 视频对象 视频对象平面 视频检索 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]