计算机技术与发展2012，Vol.22Issue(7)：60-63,4.
基于改进的混合高斯模型的目标检测方法
Target Detection Method Based on Improved Gaussian Mixture Model
摘要
Abstract
Gauss mixture model is an effective method for environmental monitoring,which can delect the abnormal changes in the environment timely,the basic thought of Gauss mixture model is when the environment and the established model matches,it is as the background, or the foreground-By using Gaussian mixture model for target detection in the image, foreground is mistakenly substituted into the model easily when model replaces,effecting background difference results. At the same time using pixels to prospect judgment,the information is often insufficient .which can not effectively express environmental characteristics. Gauss mixture model also exists the problem mat the ability of illumination effect removing is not ideal, when light occurs,the background will be changed,if not promptly detected the changes, excluding the light,it will cause false detection .affecting the accuracy of the system. Aiming at these problems,put forward a solution and use experiment to prove the validity in the paper.关键词
混合高斯模型/目标检测/背景差分/光照影响Key words
Gaussian mixture model j target detection [background differertce/illuinination effect分类
信息技术与安全科学引用本文复制引用
曙光,张超,蔡则苏..基于改进的混合高斯模型的目标检测方法[J].计算机技术与发展,2012,22(7):60-63,4.基金项目
国家自然科学基金(61075076) （61075076）
