摘要
提出一种可伸缩式镜头算法模型,利用视频镜头检测的方法,将大型视频分为背景特征变化较小的短视频。该方法使得多维缩放算法得到更精确的匹配序列,并对视频镜头进行有目的的匹配,最终达到降低大型视频匹配算法中时间复杂度的目的。实验结果表明,该算法计算量及匹配误差率明显降低。
This paper introduces the model of telescopic-shot, and uses the method of video shot detection, and separate the large video into short videos, to get the accurate matching sequences. Finally reduces the time complexity of large video. The experimental results show, the time complexity of this paper is significantly reduced, and the rate of mistake is lower.
出处
《电子技术应用》
北大核心
2012年第1期112-114,117,共4页
Application of Electronic Technique
基金
广西科学基金项目(桂科自0991058)
广西高校人才小高地建设创新团队资助计划项目(桂教人[2007]71号)
关键词
可伸缩
镜头分割
视频匹配
telescopic
shot-cutting
video matching
