基于事件流的新闻视频场景分割方法
News Video Scene Segmentation Based on Events Stream
-
摘要: 在视频数据库研究领域,视频场景的分割是一个重要而又难以解决的问题.文中从分析新闻视频场景所特有的结构特征入手,提出一种基于事件流的新闻视频场景分割方法.实验表明,采用该方法场景分割正确率可达86.9%.Abstract: For creating video database,segmentation of video scene has been an important problem hard to solve.This paper analyzes the structural features of news video and introduces a new method for scene segmentation based on events stream.Experimentation results reveal that the rate of correct segmentation is up to 86.9%.
下载: