摘要
奈奎斯特采样定理对地震数据的采集提出了很高的要求,而地震勘探的复杂环境可能导致地震数据道缺失或者勘探成本上升,因此需要对地震数据进行重建,恢复地震数据的全貌。针对上述情况,对道缺失地震数据进行双树复小波变换,使用缺失地震数据的采样矩阵作为测量矩阵,在重建阶段提出一种基于双树复小波变换的双阈值迭代重建方法。与传统的基于小波变换或基于傅里叶变换和单阈值迭代算法(IST)相比,双树复小波变换有着平移不变、多方向选择等优良特性,而使用双阈值迭代能更好的重建地震数据。实验阶段采用marmousi数据,验证了算法的可行性与有效性。
The Nyquist＇s theorem brings forward a higher request to the sampling of seismic data.While there is a big chance that the complex environment of seismic prospecting causes losing of the trace of the seismic data or increasing the cost.This is the reason why data reconstruction is needed for getting the full view of the seismic data.First,we use the dual-tree complex wavelet（DTCW）transform to process the irregular seismic data,then the sampling matrix of irregular seismic data is treated as the measurement matrix.After that,the seismic data are reconstructed by the bivariate shrinkage iteration.Unlike conventional reconstruction algorithms based on wavelet transform,Fourier transform or Iterative-Shrinkage-Thresholding（IST）algorithm,the DTCW is multi-directional and translation invariant,and the dual-threshold iteration could lead to a better reconstruction result.Experiments on the marmousi model data have verified the feasibility and efficiency of the method.
作者
唐国维
程鑫华
张岩
TANG Guowei;CHENG Xinhua;ZHANG Yan(School of Computer and Information Technology,North Eastern Petroleum University,Daqing 163318)
出处
《微型电脑应用》
2018年第11期140-144,共5页
Microcomputer Applications
关键词
地震数据重建
稀疏表示
压缩感知
迭代阈值收缩
Seismic data reconstruction
Sparse representation
Compressed sensing
Iterative shrinkage threshold
