| 快速独立分量变换与去噪初探 |
| |
| 引用本文： | 刘喜武,刘洪,李幼铭.快速独立分量变换与去噪初探[J].中国科学院研究生院学报,2003,20(4):488-492. |
| |
| 作者姓名： | 刘喜武 刘洪 李幼铭 |
| |
| 作者单位： | 1. 中国科学院地质与地球物理研究所,北京,100029;中国科学院研究生院,北京,100039
2. 中国科学院地质与地球物理研究所,北京,100029 |
| |
| 基金项目： | 中国科学院知识创新工程重大项目 (KZCXL SW 18)资助 |
| |
| 摘 要： | 独立分量分析 (ICA)通过对非高斯分布数据进行有效表示 ,获得在统计学上独立的各个分量。这种表示可以获取数据的基本结构 ,包括特征提取和信号分离。简述ICA基本理论和快速算法 ,对照主分量分析 (PCA)的Karhunen Loeve(K L)变换 ,提出独立分量变换 (ICT)概念。在分析地震信号特点的基础上 ,对模拟和实际含噪地震道进行独立分量变换、信噪分离和去噪处理初步探索 ,重建获得令人满意的去噪结果。研究表明 ,ICA在勘探地震信号处理中具有应用前景
|
| 关 键 词： | 独立分量分析 快速算法 变换 地震信号去噪 初探 |
| 修稿时间： | 2002年6月19日 |
Independent Component Transformation and its Testing Application on Seismic Noise Elimination |
| |
| LIU Xi-Wu ,LIU Hong LI You-Ming.Independent Component Transformation and its Testing Application on Seismic Noise Elimination[J].Journal of the Graduate School of the Chinese Academy of Sciences,2003,20(4):488-492. |
| |
| Authors: | LIU Xi-Wu LIU Hong LI You-Ming |
| |
| Institution: | LIU Xi-Wu 1,2LIU Hong 1LI You-Ming 1 |
| |
| Abstract: | ICA is a novel statistical method developed recently, which is used to find a representation of the Non-Gaussian multivariate data. The representation shows that each component of the computed vector is independent statistically, or as independent as possible. In application, this kind of transformation aims to capture the basic structures of the analyzed data, including features abstraction and separation of signals. Presents the fundamental theory and fast algorithms, at the same time, implement the FastICA and its updated version. Compared with PCA or K-L transformation, proposes the concept of Independent component transformation (ICT). On the basis of analyzing the features of seismic signals, does preliminary studies and try to apply ICA on seismic signal processing. Research results show the good perspective of ICA application to seismic signal processing. |
| |
| Keywords: | independent component analysis fast algorithm transformation seismic signal noise elimination attempt |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
