摘要
镜头分割是基于内容的视频检索(CBVR)的关键步骤,它会直接影响到视频检索的结果。针对传统镜头分割算法实现过程复杂,渐变检测不理想等不足,提出了颜色直方图变化率HCR(Histogram Change Ratio)的概念,并在此基础上提出了一种基于颜色直方图变换率的视频镜头分割算法。实验证明,该镜头分割算法实现过程简单,而且对镜头突变和渐变均具有很好的检测效果。
Video shot segmentation is the key procedure of content-based video retrieval（CBVR）,it affects the results of video retrieval directly.Traditional shot boundary segmentation algorithms are complex in its implementation process,and are not suitable for gradual change detection.To overcome this problem,in this paper we make a definition on colour histogram change ratio（HCR） and then present a video shot segmentation approach based on colour HCR.As proved by the experimental results,the approach is brief and effective for abrupt change and gradual change detection.
出处
《计算机应用与软件》
CSCD
2011年第4期17-20,共4页
Computer Applications and Software
基金
国家自然科学基金项目(60772122)
中国高等教育博士项目研究基金(20070357001)
安徽省高校优秀青年人才项目(2009SQRZ019ZD
2009SQRZ221)
关键词
颜色直方图变化率
镜头分割
渐变
Colour histogram change ratio Shot boundary segmentation Gradual change
