一种改进的背景提取算法  

Modified Background Extraction Algorithm

作　　者：张康宁[1] 赵秋娣[1] 李祥艳[1] 张天行[1] 

机构地区：[1]长春理工大学电子信息工程学院,长春130022

出　　处：《计算机系统应用》2014年第7期144-147,共4页Computer Systems & Applications

摘　　要：背景提取,是利用背景差分法获取目标的重要步骤.针对传统背景提取算法需要处理大数据量的问题,本文算法在传统背景提取算法的基础上,利用一段时间内,相邻图像相同位置的灰度值基本处在同一平稳区间的原理和思路,通过对小数据量的多次操作来获取背景图像.实验结果证明,这种做法可以在使用较少内存使用量的基础上获得比较好的效果。Background extraction is one of the most important steps to get the obtain target by using background difference. The question of the traditional background extraction is to deal with the large amount of data. In order to deal with this problem, this paper proposes a modified background extraction algorithm, based on the traditional background extraction algorithm. It is the principle of the algorithm that the grey value of adjacent images and the same position belongs to the same smooth range in a period of time. According to the principle, we can deal with small amount of data by deal with many groups. The experimental shows that this approach can be getting a better effect, and using less memory.

关 键 词：目标检测 背景提取 内存使用量 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]