【摘要】：在地震數(shù)據(jù)處理流程中,通常對不規(guī)則的、稀疏的或者缺失的地震數(shù)據(jù)進(jìn)行插值處理,通過插值方法來避免多次波的預(yù)測錯誤和成像假頻等現(xiàn)象,使地震數(shù)據(jù)處理更加精準(zhǔn)。Shearlet變換是一種多尺度變換,具有最佳的稀疏性、方向性以及局部化特性。將Shearlet變換與基于Landweber加速下降迭代方法結(jié)合起來對地震數(shù)據(jù)進(jìn)行插值,在保證求解精度的同時提高了計算效率。信號和噪聲在Shearlet域具有不同的分布特點,通過閾值法壓制隨機噪聲,可提高算法的抗噪性。此外,采用jitter采樣的方式,更好地壓制了假頻信息。理論和實際地震數(shù)據(jù)驗證了該方法的有效性。
[Abstract]:In the process of seismic data processing, irregular, sparse or missing seismic data are usually interpolated to avoid multiple prediction errors and false imaging frequency. Making the seismic data processing more accurate, the Shearlet transform is a multi-scale transform with the best sparsity, directionality and localization characteristics. The Shearlet transform is combined with the accelerated descent iteration method based on Landweber to interpolate the seismic data, which not only ensures the accuracy of the solution, but also improves the computational efficiency. The signal and noise have different distribution characteristics in Shearlet domain. The anti-noise of the algorithm can be improved by using threshold method to suppress random noise. In addition, jitter sampling is used to suppress the false frequency information better. The validity of the method is verified by theoretical and practical seismic data.
【作者單位】： 吉林大學(xué)地球探測科學(xué)與技術(shù)學(xué)院;
【基金】：國家科技重大專項項目(2011ZX05023-005-008) 國家自然科學(xué)基金項目(41374108)~~
【分類號】：P631.44
，

本文編號：2170585