Time-frequency characterization of seismic data using local attributes
Time-frequency characterization of seismic data using local attributes
复制标题
DOI：
10.1190/1.3255210
复制
发表时间：
2009
期刊：
影响因子：
--
通讯作者：
Guochang Liu;Sergey Fomel;Xiaohong Chen
中科院分区：
文献类型：
--
作者：
Guochang Liu;Sergey Fomel;Xiaohong Chen
地震数据作为一种非平稳信号，在时间上可能具有不同的频率。为了局部表示频率分量，我们提出了一种新的计算非平稳频率特征的方法，该方法使用迭代求逆框架中计算的非平稳傅立叶系数。我们不使用滑动窗口，而是使用整形正则化来控制表征的局部性。根据局部频率特征图可以估计地震数据的局部主频。局部频率特征的自适应计算是该方法的一个优点，合成和现场数据实例证明了这一点。
Seismic data, as a nonstationary signal, can have varying frequency in time. To locally represent frequency components, we present a novel method of computing nonstationary frequency characterization using nonstationary Fourier coefficients, which are computed in an iterative inversion framework. Instead of using sliding windows, we use shaping regularization to control locality of the characterization. The local dominant frequency of seismic data can be estimated from a local frequency characterization map. Adaptive computation of local frequency characterization is an advantage of our method, which is demonstrated by synthetic and field data examples.