基于颜色特征的视频数据库检索系统

为了在视频数据库中提供有效的视频检索和浏览功能,必须建立高效的索引.由于视频数据具有层次性的结构,在镜头边界检测后,可以利用聚类方法按不同的相似性尺度对镜头关键帧进行处理,对视频数据建立索引.该系统采用颜色特征,使用Twin Comparison算法实现镜头检测和直方图平均法实现关键帧提取,对关键帧采用K均值聚类算法处理,建立视频数据库索引.实验结果表明该系统能较好地实现视频快速浏览和检索功能.

Video database retrieval system based on color information

To provide effective video browse and retrieval ability, high-powered index is designed. Because of the hierarchical structure of video data, after shot boundary detection, cluster method with different similarity metrics is used to deal with key frame of shot and to establish video data indexing. Using color feature, twin comparison algorithm is adopted to detect shot and mean color histogram to extract key frame. And k-means clustering algorithm is adopted to establish data indexing. The experimental show that result this system make video retrieval much easier.