MCA框架下K-SVD构建字典对地震信号去噪与重建  

Denoising and Reconstruction of Seismic Signals based on K-SVD under the MCA Framework

作　　者：陈思琪 薛雅娟[1] 杨清蜜 方立鑫 郑书琳 CHEN Siqi;XUE Yajuan;YANG Qingmi;FANG Lixin;ZHENG Shuling(College of Communication Engineering,Chengdu University of Information Technology,Chengdu 610225,China;Sinopec Key Laboratory of Geophysics,Nanjing 21000,China)

机构地区：[1]成都信息工程大学通信工程学院,四川成都610225 [2]中国石化地球物理重点实验室,江苏南京210000

出　　处：《成都信息工程大学学报》2021年第1期7-14,共8页Journal of Chengdu University of Information Technology

基　　金：四川省杰出青年学术技术带头人资助项目(2016JQ0012)。

摘　　要：针对由于地震信号采集环境的复杂性带来的采样信号不完整,存在大量噪声等情况,对采集的叠前地震信号进行去噪和重建。在传统的K-奇异值分解(K-singular value decomposition,K-SVD)建造冗余字典的压缩感知重建的基础上,提出了基于形态分量分析(morphometric principal components analysis,MCA)的K-SVD地震信号的去噪与重建。即使用MCA对地震信号的结构和平滑部分进行分类,并针对上述两种类别分别构建由K-SVD算法计算的冗余字典,将两种类别分别置于不同字典中进行去噪与重建。与传统的方法相比,该方法在减少了地震信号采集的成本和难度的基础上,精确辨别地震信号细节,并取得良好的去噪效果。For the incompleteness of sample signal due to the complexity of seismic signal collection environment,and quantities of noise.Therefore,before analyzing it,we need to denoise and reconstruct the pre-stack seismic data.On the basis of building the compress sensing reconstruction of redundant dictionaries by traditional K-singular value decomposition(K-SVD),we propose the K-SVD seismic signal denoise and reconstruction combined with moqihometric principal components analysis(MCA).It realizes that we use MCA to classify the structure and smooth part of the seismic signal.And build redundant dictionaries calculated by the K-singular value decomposition for the above two parts,respectively.Finally,the two parts are placed in different dictionaries for denoising and reconstruction.Compared with the traditional method,this method reduces the cost and difficulty of seismic data acquisition,accurately discriminates signal details,and achieves good denoising effect.

关 键 词：地震信号 压缩感知 K-奇异值分解 形态分量分析 冗余字典 

分 类 号：TN911.4[电子电信—通信与信息系统]