摘要
提出了一种新的相似视频快速检索方法.根据视频的时空分布统计得到图像特征码和视频单元,通过统计视频单元数量度量视频相似性.为了适应可扩展计算的需要,提出了基于聚类索引表的检索方法.通过对大规模数据库的查询测试证明该相似性检索算法快速有效.
A novel fast video similarity search approach was proposed. Image characteristic code and video component were computed based on the statistics of video spatial-temporal distribution. Video similarity was measured by computing the number of video components. An efficient search method based on clustering index table was proposed for the requirement of scalable computing. The query test results from large video database show that the proposed video similarity search algorithm is efficient and effective.
出处
《中国科学院研究生院学报》
CAS
CSCD
北大核心
2010年第3期376-380,共5页
Journal of the Graduate School of the Chinese Academy of Sciences
基金
国家863计划项目(2008AA01Z408)
国家自然科学基金(60975045)资助
关键词
相似视频检索
图像特征码
聚类索引表
时空分布特征
video similarity search
image characteristic code
clustering index table
spatial-temporal distribution
