摘要
在借鉴了计算机视觉技术和多媒体处理算法的基础上,提出了一种基于全自动高效视频分析框架的分析和总结的体育视频分析结构。本视频分析框架拟从底层特征提取、中级关键基元生成出发,初步实现体育视频中的镜头分类和运动对象识别以及跟踪等。分析框架中包含了一些中低级别的足球视频镜头分类处理算法,如主色区域检测,鲁棒镜头边界检测,以及目标检测等等。
On the basis of computer visual technology and multimedia processing algorithm, the paper proposes a sports video analysis structure on the basis of analysing the automatic efficient video analysis framework. The video analysis frame work starts from extracting the bottom features and intermediate key element generation and conducts the shot classification, moving object identification and tracking. The analysing framework contains some shot classification process algorithm of the football video, such as color region detection, robust lens boundary detection, and target detection, etc.
出处
《体育成人教育学刊》
2012年第3期49-51,共3页
Journal of Sports Adult Education
基金
中国博士后科学基金第五十批面上资助(2011M501369)
高校基本科研业务费中山大学文科青年教师培育项目(12WKPY53)
关键词
体育视频分析
镜头分类
事件检测
特征提取
sports video analysis
shot classification
event detection
feature extraction
