多目标视频运动分析系统中的多几何特征参数快速提取实现  

Fast geometric parameters extraction in multitargetvideo motion processing systems

作　　者：吴四九[1] 蔺冰[1] 

机构地区：[1]成都信息工程学院,四川成都610041

出　　处：《成都信息工程学院学报》2002年第4期246-250,共5页Journal of Chengdu University of Information Technology

摘　　要：首先通过将整个图像区域看成系列行、行内的各线段构成的集合,然后判断相邻行各线段是否满足合并条件,最后根据判断结果实现对应的目标区域几何特征参数的统计。理论与应用都表明该算法优点是:对图像只扫描一次,即可同时、并行地获得多目标区域多个几何特征参数。Firstly, the image is decomposed into many lines, each of which consists of a series of short line segments. Secondly, the different short line segments in the adjacent lines are compared to see whether the region connection condition is satisfied. Finally, the target geometric parameters are estimated. Both the theory and application show that this algorithm needs only one scanning for the image to extract all the geometric parameters concurrently.

关 键 词：多目标视频运动分析系统 多几何特征参数 几何特征描述 实时性分析 快速参数检测 图像处理 

分 类 号：TN911.73[电子电信—通信与信息系统]