摘要
关键帧提取是基于内容检索的一个重要的组成部分,所提取的关键帧的质量的好坏,直接影响检索的结果。该文介绍了目前几种较为典型的算法,并在对其进行分析比较的基础上,针对新闻视频提出了基于文字和图像信息提取关键帧的算法,取得了很好的效果,最后在结论中提出了综合运用音频、图像、文字和运动信息提取关键帧的层次化算法思想。
Key frame is the frame which can represent the salient content of the shot.Depending on the content complexity of the shot,one or more key frame can be extracted from a single shot.Key frame extraction has been recognized as one of the important research issues in content-based retrieval.This paper presents some typical approach to extracting key frame from video,and analyzes their difference and shortcomings.Then the paper proposes a algorithm based on text and image information to extract key frame from news video,and the effectiveness has been validated by experiments.
出处
《计算机工程与应用》
CSCD
北大核心
2002年第9期83-85,共3页
Computer Engineering and Applications
基金
国家电力公司计算机应用重点学科专项基金
