| KECCA方法及其在水下回波分类中的应用 |
| |
| 引用本文： | 刘志勇,夏毅敏,杨勃. KECCA方法及其在水下回波分类中的应用[J]. 郑州大学学报(工学版), 2011, 0(4) |
| |
| 作者姓名： | 刘志勇 夏毅敏 杨勃 |
| |
| 作者单位： | 中南大学机电工程学院;桂林航天工业高等专科学校; |
| |
| 基金项目： | 国家自然科学基金资助项目(50875265); 中国博士后科学基金资助项目(20080440992); 湖南省科技支撑计划(2009SK3159) |
| |
| 摘 要： | 针对增强典型相关方法不能提取对象的非线性特征的问题,在增强典型相关方法基础上,提出了核增强典型相关方法.该方法主要用于非线性相关关系分析,能够提取出类内非线性相关度最大且类间非线性相关度最小的多特征融合信息.最后,将基于核增强典型相关的多特征融合方法应用到水下回波分类领域.4种水下底质回波分类实验表明,与两种单特征相比,采用核增强典型相关多特征融合方法,平均分类正确率提高了3%,验证了基于核增强典型相关的多特征非线性融合的有效性.
|
| 关 键 词： | 特征级融合 核增强典型相关方法 水下回波分类 |
KECCA Method and Its Application in Underwater Echoes Classification |
| |
| LIU Zhi-yong,,XIA Yi-min,YANG Bo. KECCA Method and Its Application in Underwater Echoes Classification[J]. Journal of Zhengzhou University: Eng Sci, 2011, 0(4) |
| |
| Authors: | LIU Zhi-yong XIA Yi-min YANG Bo |
| |
| Affiliation: | LIU Zhi-yong1,2,XIA Yi-min1,YANG Bo1(1.School of Mechanical and Electrical Engineering,South Central University,Changsha 410083,China,2.Guilin College of Aerospace Technology,Guilin 5410043,China) |
| |
| Abstract: | Enhanced canonical correlation analysis(ECCA) cannot Abstract objects' nonlinear features.An kernel enhanced canonical correlation analysis(KECCA) is proposed based on ECCA.KECCA is mainly used to analyze the nonlinear relationship of two multiple dimentional datasets,and using it the fusion information of the most within-class nonlinear relation and the least between-class nonlinear relation can be Abstracted.Finally,the fusion model based on KECCA is used on underwater echo classification and the experime... |
| |
| Keywords: | feature level fusion KECCA underwater echo classification |
| 本文献已被 CNKI 等数据库收录！ |
|
