| 基于高斯混合模型的铁路入侵物体目标识别方法 |
| |
| 引用本文： | 董宏辉,孙智源,葛大伟,秦勇,贾利民.基于高斯混合模型的铁路入侵物体目标识别方法[J].中国铁道科学,2011,32(2). |
| |
| 作者姓名： | 董宏辉 孙智源 葛大伟 秦勇 贾利民 |
| |
| 作者单位： | 北京交通大学轨道交通控制与安全国家重点实验室,北京,100044 |
| |
| 基金项目： | 国家“八六三”计划项目(2006AA11Z231); 铁道部科技研究开发计划项目(2007X007-A); 北京市科技计划重点项目(D07020601400707) |
| |
| 摘 要： | 基于铁路的特殊视频场景,研究铁路入侵物体目标识别技术,提出并应用1种改进的高斯混合模型,拟定合适的颜色变化阈值和背景更新速率,对图像的不同部分分别进行背景更新,实现铁路视频的背景建模,并由此得到稳定的背景图像;通过背景像素与前景像素的贝叶斯分类实现对铁路入侵物体的准确检测。对典型的铁路入侵行为视频进行实验分析,结果表明:应用改进的高斯混合模型可以更好地适应场景的变化,并能够更加快速、准确地实现在铁路环境下对入侵物体的目标识别。
|
| 关 键 词： | 高斯混合模型 背景建模 运动检测 目标识别 铁路安全 |
Target Recognition Method of Railway Invasion Based on Gaussian Mixture Model |
| |
| DONG Honghui,SUN Zhiyuan,GE Dawei,QIN Yong,JIA Limin.Target Recognition Method of Railway Invasion Based on Gaussian Mixture Model[J].China Railway Science,2011,32(2). |
| |
| Authors: | DONG Honghui SUN Zhiyuan GE Dawei QIN Yong JIA Limin |
| |
| Institution: | DONG Honghui,SUN Zhiyuan,GE Dawei,QIN Yong,JIA Limin(State Key Laboratory of Rail Traffic Control and Safety,Beijing Jiaotong University,Beijing100044,China) |
| |
| Abstract: | Based on the specific video of railway scene,the target recognition of railway invasion is studied,and an improved Gaussian Mixture Model(GMM) is proposed in this paper.This improved GMM can get more stable backgrounds,by setting the different values of the color changing threshold and background updating rate,and by updating the different image parts,respectively.Using the Bayesian classification for the background pixel and foreground pixel,the railway invasion is detected accurately.The experiment result... |
| |
| Keywords: | Gaussian mixture model Background modeling Motion detection Target recognition Railway safety |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
