摘要
镜头切换检测是视频检索研究领域中一项基本的、重要的技术.然而传统的镜头边界检测方法都只考虑了视频相邻两帧的信息,没有充分考虑到视频序列中前后的连续性,从而容易导致误检测.基于此,提出了一种基于类间距离和准则对镜头边界进行检测的新方法,该方法能够充分利用视频序列中的连续性信息,克服传统镜头边界检测方法的不足,有效地提高了检出率及精确率.大量实验表明,该方法是行之有效的.
The shot change detection is a basic and important technology in video retrieval. But the traditional shot boundary detection methods only consider the two adjacent video information, and do not take fully into account the pre- vious and consequent video sequences in the continuity and thus easily lead to error detection. Therefore a new shot boundary detection algorithm based on maximum-between-cluster-distance is proposed. The method can make full use of video sequenees in the continuity of information, and overcome the traditional boundary-less than the detection method. Large number of experiments shows that the algorithm has effectively improved the detection rate and precision.
出处
《南京师范大学学报（工程技术版）》
CAS
2008年第4期108-111,共4页
Journal of Nanjing Normal University(Engineering and Technology Edition)
关键词
视频检索
镜头分割
类间距离
video detection, shot boundary detection, class distance
