河道监控视频的运动对象分割算法  

Moving object segmentation algorithm for river surveillance video

作　　者：叶利华[1] 殷海明[1] 魏远旺[1] 

机构地区：[1]嘉兴学院数理与信息工程学院,浙江嘉兴314001

出　　处：《计算机应用》2011年第1期263-265,共3页journal of Computer Applications

基　　金：嘉兴市科技局计划项目(2009AY2074)

摘　　要：针对内河航道监控视频的特点,提出一种基于背景差法的对象分割算法。首先在HSI颜色空间里利用像素的色调和亮度对其进行归类;然后利用基于块处理的方法确定背景像素,并在背景缓慢变化和急速变化时,采用定时和实时的背景重构方法进行背景更新;最后利用背景差提取运动对象。An object segmentation algorithm based on background subtraction was presented according to the features of river surveillance video. First pixels were classified by the hue and intensity values in the HSI color space; then the background pixels were determined by a block-based algorithm; at last the moving objects were extracted with the background subtraction.

关 键 词：背景差 背景重构 运动检测 视频对象分割 视频监控 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]