计算机应用与软件2017，Vol.34Issue(4)：238-242,5.DOI:10.3969/j.issn.1000-386x.2017.04.040
一种基于改进混合高斯模型的前景检测
A FOREGROUD DETECTION BASED ON IMPROVED GAUSSIAN MIXTURE MODEL
摘要
Abstract
A new algorithm (TGM) for foreground detection is proposed based on improved Gaussian Mixture Model to solve the problem of huge computation of classic Gaussian Mixture Model (GGM).The quantity of Gaussian distribution of the pixels in background stability region is decreased and the computation is reduced based on the Model clean-up mechanism of historical information.Besides, the temporary Gaussian distribution is built and the easier modified operation is utilized to further reduce computation.At the last, the temporary Gaussian models which match the conditions are turned into official Gaussian models to avoid updating models meaninglessly and improve veracity.Experimental results show that the improved algorithm is feasible with better instantaneity and veracity.关键词
前景目标检测/混合高斯模型/模型清理机制/临时高斯分布Key words
Foreground detection/Gaussian Mixture Model/Model clean-up mechanism/Temporary Gaussian distribution分类
信息技术与安全科学引用本文复制引用
罗向荣,李其申..一种基于改进混合高斯模型的前景检测[J].计算机应用与软件,2017,34(4):238-242,5.基金项目
江西省自然科学基金项目(YC2014-S394). （YC2014-S394）
