该文描述了一种有效的框架对足球视频进行结构分析,根据电影特征和对象特征生成视频概要。由于足球视频的特殊性,本文在镜头边界检测中采用分层检测的方法:象素点对的比较、颜色直方图和对象分割和跟踪技术。我们在镜头分类中对中远镜头的区分提出了新的方法。以慢动作回放镜头为标志,通过分析镜头间的关联规则生成视频概要。
In this paper,we present an effective framework for soccer video analysis and summarize the soccer video by using cinematic and object features. The shot boundary detector employs the multi-filter structure:pixel level comparison,histogram,object segmentation maps and object tracking for the particularity in soccer video. We present a new method to partition long shot and medium shot in shot classification. For ever slow-motion replay,the system generates summarization by using association rule in shots.