一种改进的背景提取与更新方法  

An Improved Method on Background Extraction and Update

作　　者：周建强[1] 刘怀[1] 张海龙[1] 李振[1] 

机构地区：[1]南京师范大学电气与自动化工程学院,江苏南京210042

出　　处：《南京师范大学学报（工程技术版）》2012年第4期67-72,共6页Journal of Nanjing Normal University(Engineering and Technology Edition)

基　　金：国家自然科学基金(51075215)

摘　　要：在分析背景提取的特点、均值算法准确度低和高斯混合模型法实时性差等不足的基础上,提出了一种改进的背景提取和更新的算法.该算法通过结合概率分布和统计直方图来得到初始背景,以降低环境变化的影响;利用高斯分布和递归算法来提高区域背景更新的实时性和准确度.结果表明,与均值算法和混合高斯模型法相比,该算法实现简单,具有很好的准确性,并且抑制了目标运动过慢造成的干扰,提高了背景更新的实时性.The paper introduces an improved method on background extraction and update based on the features of background extraction and the disadvantages of the low aecurancy of mean algorithm and the bad real-time of mixture Gaussian model. The method obtains the initial background with probability distribution and statistics of histogram to reduce the impact of environmental changes. Then to increase the real-time and accurancy,it combines the Gaussian distribution with recursive algorithm to update the background. Compared with means algorithm and mixture Gaussian model,it is easier to be operated, has better real-time and aecurancy, and can effectively inhibit the noise caused by the slow targets.

关 键 词：背景提取 概率分布 统计直方图 递归算法 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]