利用改进NFL算法对镜头进行基于内容的检索  

Content-Based Retrieval of Video Shot Using the Improved Neatest Feature Line Method

作　　者：赵黎[1] 杨士强[1] 祁卫[2] 李子青[2] 张宏江[2] 

机构地区：[1]清华大学计算机科学与技术系,北京100084 [2]微软亚洲研究院,北京100080

出　　处：《软件学报》2002年第4期586-590,共5页Journal of Software

基　　金：国家重大基础研究973发展规划资助项目(G1999032704);清华大学-微软公司多媒体技术实验室资助项目

摘　　要：基于镜头的分类和检索对于视频库的管理和查询非常重要.将“最近特征线”法（nearest feature line,简称NFL）用于镜头的分类和检索.将镜头中的代表帧看做是某个特征空间中的点,通过这些点间的连线表征该镜头的总体特征信息,然后计算查询图像和特征线的距离,以决定镜头与查询图像的相似度.为了更适于视频数据,对原来的NFL方法进行了改进,基于镜头内部内容活动程度对特征线进行限制、实验结果表明,改进的NFL方法比传统的NFL方法以及常用的聚类万法,如最近邻法（nearest neighbor,简称NN）和最近中心法（nearest center,简称NC）,在性能上有所提高.The shot based classification and retrieval is very important for video database organization and access. In this paper, a new approach NFL (nearest feature line) used in shot retrieval is presented. Key-Frames in shot are looked as feature points to represent the shot in feature space. Lines connecting the feature points are further used to approximate the variations in the whole shot. The similarity between the query image and the shots in video database are measured by calculating the distance between the query image and the feature lines in feature space. To make it more suitable to video data, the original NFL method by adding constrains on the feature lines is improved. Experimental results show that the improved NFL method is better than the traditional classification methods such as the nearest neighbor (NN) and the nearest center (NC).

关 键 词：内容检索 最近特征线 视频检索 视频分类 视频镜头 NFL算法 视频数据库 

分 类 号：TP392[自动化与计算机技术—计算机应用技术]