利用单个特征识别强噪声中的弱小运动目标,常因所提取的目标特征与噪声特征易混淆而导致高的虚警率.提出一种新的基于多特征融合的弱小运动目标识别方法.分析了弱小运动目标的连续相关性、面积及质心位置偏移这三个特征的可靠性及提取方法,对获取的特征值进行归一化后采用多特征融合的方法构造更具有鲁棒性的联合特征,确定了以具有最大多特征融合值为真实目标的决策方法.通过与采用单一特征的目标识别方法进行比较,证明了提出的多特征融合方法能更准确地识别弱小运动目标.
参考文献
| [1] | Dong N L,Jin G,Qi B.New approach to detect dim moving point targets based on motion analysis[C] // SPIE,2001,4473:34-42. |
| [2] | Zhang P F,Li D S,Wang Q.A novel iris recognition method based on feature fusion[C] // Proc.of Third International Conference on Machine Learning and Cybernetics,2004,3661-3665. |
| [3] | Chu S M,Libal V,Marcheret E,et al.Multistage information fusion for audio-viaual speech recognition[C] //Proc.IEEE IEMC,2004,1651-1654. |
| [4] | Zheng Lin,Han Chongzhao,Zuo Dongguang,et al.Moving object recognition based on multi-feature fusion[J].Journal of System Simulation (系统仿真学报),2004,16(5):1081-1084 (in Chinese). |
| [5] | Wu Wei,Peng Jiaxiong.The feature of small target and its invariance analysis in infrared image sequence[J].Journal of Huazhong University of Science and Technology (华中科技大学学报),2002,30(3):83-85 (in Chinese). |
上一张
下一张
上一张
下一张
计量
- 下载量()
- 访问量()
文章评分
- 您的评分：
-
10%
-
20%
-
30%
-
40%
-
50%