摘要
针对现有的背景重构算法存在的问题,提出一种新的算法,该算法基于子块操作,利用对称差分重构背景。首先对对称差分算法进行了改进,然后基于对称差分结果进行背景重构。该算法一定程度上利用了相邻像素间的关系,计算简单,不需存储采样图像,且对光照和背景变化可以自适应更新。实验结果表明,该算法能够快速、有效地重构背景图像。
A new background reconstruction algorithm is proposed to overcome the problems in the current algorithms,which is based on sub-blocks and employs symmetrical differencing.First,the original symmetrical differencing is improved.Then,background is reconstructed based on the symmetrical differencing results.This algorithm makes use of the spatial information in part,only need simple computation,and need not store history images.Moreover,it can self-adapt to the changes of illumination and of background.Experimental results indicate the fast and efficient performance of the algorithm in background reconstruction.
出处
《计算机工程与应用》
CSCD
北大核心
2008年第6期104-106,共3页
Computer Engineering and Applications
基金
国家自然科学基金(the National Natural Science Foundation of Chinaunder Grant No.60104009)
山东省自然科学基金(the Natural Science Foundation of Shandong Province of Chinaunder Grant No.Z2005G03)。
关键词
对称差分
背景重构
子块
symmetrical differencing
background reconstruction
sub-block
