| 鲁棒性自学习背景建模方法研究 |
| |
| 引用本文： | 姜华,刘辉,刘鑫,王伟嘉. 鲁棒性自学习背景建模方法研究[J]. 山西电子技术, 2008, 0(3): 88-90 |
| |
| 作者姓名： | 姜华 刘辉 刘鑫 王伟嘉 |
| |
| 作者单位： | 昆明理工大学信息工程与自动化学院,云南昆明650051 |
| |
| 摘 要： | 在利用混合高斯模型构建动态背景的基础上,提出了一种抑制错误的前景检测,克服无关结构干扰的前景/背景分割方法,并利用颜色信息去除图像中的阴影部分。多个复杂室外场景下的实验结果表明,该方法能够克服背景晃动和摇晃树枝干扰,并能够抑制光影变化（影子或强光）的影响,具有较好的鲁棒性。
|
| 关 键 词： | 混合高斯 背景差分 误判检测 阴影检测 |
Self-Learning Method of Robust Background Modeling |
| |
| Jiang Hua Liu Hui Liu Xin Wang Wei-jia. Self-Learning Method of Robust Background Modeling[J]. Shanxi Electronic Technology, 2008, 0(3): 88-90 |
| |
| Authors: | Jiang Hua Liu Hui Liu Xin Wang Wei-jia |
| |
| Affiliation: | Jiang Hua Liu Hui Liu Xin Wang Wei-jia (Department of Information , Automatic,Kun Ming University of Science , Technology,Kunming Yunnan 650051,China) |
| |
| Abstract: | In this paper it introduces a foreground/background segmentation approach that can suppress the false detection due to noise and other unrelated structures based on mixture Gaussian background modeling.The approach can suppress shadows of the targets from being detected due to illumination changes also.The experiments run in outdoor environment indicates that the background model can handle situations where the background of the scene is not completely static but contains small motions such as tree branch m... |
| |
| Keywords: | mixture Gaussians background subtraction false detection shadow detection |
| 本文献已被 CNKI 维普 等数据库收录！ |
|
