摘要
为了实现视频层次结构挖掘,提出了一个有效的视频镜头分割算法和一种镜头相似性度量方法,然后根据镜头颜色、纹理和语义相似性采用场景边界探测算法构造视频场景。实验结果表明,这些算法是可行的。
To achieve such a hierarchical structure mining, this paper proposed an efficient shot detection algorithm, developed a novel algorithm for measuring video shot similarity and adopt Scene Change Detection (SCD) to construct video scene based on color, texture and semantic similarity between shots, The experimental results demonstrate the performance of the algorithms is feasible,
出处
《计算机应用研究》
CSCD
北大核心
2007年第6期155-157,160,共4页
Application Research of Computers
基金
国家自然科学基金资助项目(60273066)
普通高等学校教学改革项目(湘教通[2005]280号文件)
关键词
视频层次结构
视频结构挖掘
场景边界检测
video hierarchical structure
video structure mining
scene change detection