Seismic deblending by sparse inversion over dictionary learning

By dictionary learning from trained samples, a set of datadriven dictionary atoms can be obtained and used to favor sparser representation of patched data. In this paper, in the context of the blending framework with time-dithering sequential source shooting and by regarding seismic deblending as an inverse problem, sparse inversion and dictionary learning are combined to construct the corresponding formula of minimization with respect to unknown recovery, dictionary, and coefficient set. And an alternating algorithm is presented and in each iteration, dictionary atoms and coefficients are updated by K singular vector decomposition (K-SVD) method in one step and subsequently the recovery is updated by a fast steepest descent gradient method in the other step. A synthetic and real field data demonstrate the effectiveness of our method. And the outcome can be a significant reference in designing highefficiency and low-cost blending acquisition.

[1]  G. Hampson,et al.  Acquisition Using Simultaneous Sources , 2007 .

[2]  Gerrit Blacquière,et al.  From Simultaneous Shooting to Blended Acquisition , 2008 .

[3]  Ed Kragh,et al.  A new simultaneous source separation algorithm using frequency-diverse filtering , 2012 .

[4]  Gary Hampson,et al.  Acquisition using simultaneous sources , 2008 .

[5]  Yi Luo,et al.  Simultaneous sources separation via multidirectional vector-median filtering , 2012 .

[6]  Pascal Frossard,et al.  Dictionary learning: What is the right representation for my signal? , 2011 .

[7]  Youli Quan,et al.  Image integration with learned dictionaries and application to seismic monitoring , 2011 .

[8]  James Rickett,et al.  Simultaneous Source Separation By Sparse Radon Transform , 2008 .

[9]  Rajat Raina,et al.  Efficient sparse coding algorithms , 2006, NIPS.

[10]  M. Elad,et al.  $rm K$-SVD: An Algorithm for Designing Overcomplete Dictionaries for Sparse Representation , 2006, IEEE Transactions on Signal Processing.

[11]  Gerrit Blacquière,et al.  Separation of blended data by iterative estimation and subtraction of blending interference noise , 2011 .

[12]  Yoram Bresler,et al.  MR Image Reconstruction From Highly Undersampled k-Space Data by Dictionary Learning , 2011, IEEE Transactions on Medical Imaging.

[13]  I. Moore,et al.  Simultaneous Sources – Processing and Applications , 2010 .

[14]  Jianwei Ma,et al.  Seismic data denoising based on learning-type overcomplete dictionaries , 2012, Applied Geophysics.

[15]  Craig J. Beasley,et al.  A new look at marine simultaneous sources , 2008 .

[16]  Felix J. Herrmann,et al.  Randomized marine acquisition with compressive sampling matrices , 2012 .

[17]  Laurent Demanet,et al.  Fast Discrete Curvelet Transforms , 2006, Multiscale Model. Simul..

[18]  Mauricio D. Sacchi,et al.  Sparse Coding For Data-driven Coherent And Incoherent Noise Attenuation , 2009 .

[19]  Gerrit Blacquière,et al.  Separation of Blended Impulsive Sources using an Iterative Approach , 2010 .

[20]  R. Abma,et al.  High Quality Separation of Simultaneous Sources by Sparse Inversion , 2010 .