| 基于均值漂移的背景建模方法 |
| |
| 引用本文： | 杜会斌,张瑞,蒋国峰. 基于均值漂移的背景建模方法[J]. 无线电工程, 2011, 41(7): 42-44,50 |
| |
| 作者姓名： | 杜会斌 张瑞 蒋国峰 |
| |
| 作者单位： | 1. 中国电子科技集团公司第五十四研究所,河北石家庄,050081
2. 海军装备部,北京,100841 |
| |
| 摘 要： | 介绍了一种从包含复杂运动场景的图像序列中获取理想背景模型的有效方法。通过分析背景建模的若干影响因素为理想背景构建模型。使用非参数化的均值漂移算法寻找背景图像像素值分布最为集中的区域,并且实现背景模型的不断更新。实验结果表明,该方法较传统背景生成算法具有较强的鲁棒性,能够有效实现在复杂场景下的背景建模和目标检测。
|
| 关 键 词： | 运动目标检测 背景建模 背景减除 均值漂移 |
Research on Background Generation Based on Mean-shift |
| |
| DU Hui-bin,ZHANG Rui,JIANG Guo-feng. Research on Background Generation Based on Mean-shift[J]. Radio Engineering of China, 2011, 41(7): 42-44,50 |
| |
| Authors: | DU Hui-bin ZHANG Rui JIANG Guo-feng |
| |
| Affiliation: | DU Hui-bin1,ZHANG Rui1,JIANG Guo-feng2(1.The 54th Research Institute of CETC,Shijiazhuang Hebei 050081,China,2.Naval Equipment Department,Beijing 100841,China) |
| |
| Abstract: | The article presents a method for generating background model from a sequence of images with complex scenes.We introduce a model to describe the variation of the background after analyzing several background effect components.We can get the best estimate of the background pixel underlying distribution and update our model using nonparametric mean-shift approach.Experimental results from real sequences demonstrate the effectiveness of the method in background generation and object detection. |
| |
| Keywords: | moving object detection background generation background subtraction mean-shift |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
