华中科技大学学报（自然科学版）2011，Vol.39Issue(z2)：378-381,4.
基于充分统计量的在线背景减除方法
Online background subtraction based on sufficient statistics
摘要
Abstract
New online k-means algorithm based on sufficient statistic and Gaussians mixture model (GMM) for real-time segmentation of background subtraction was proposed. This algorithm, with software assignment of k-means instead of hardware assignment in EM algorithm, can advance the computational speed and the convergence rate. Experiments show that the online k-means algorithm could effectively detect moving objects, improves the computation and convergence rate, and requires less space.关键词
在线算法/收敛性/高斯混合/背景减除/充分统计量/运动目标Key words
online algorithm/ convergence/ Gaussian mixture/ background subtraction/ sufficient statistics/ moving objects分类
信息技术与安全科学引用本文复制引用
袁舟,钱徽,陈鹏,朱淼良..基于充分统计量的在线背景减除方法[J].华中科技大学学报（自然科学版）,2011,39(z2):378-381,4.基金项目
国家自然科学基金资助项目(90820306). （90820306）
