Image enhancement by multi-parameter characterization of common image gathers

Raanan Dafni*, Moshe Reshef

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

Abstract

We present an innovative approach for seismic image enhancement using multi-parameter angle-domain characterization of common image gathers. A special subsurface angle-domain imaging system is used to generate the multi-parameter common image gathers in a summation-free image space. The imaged data associated with each common image gathers depth point contain direction-dependent opening-angle image contributions from all the available incident and scattered wave-pairs at this point. Each direction-dependent opening-angle data can be differently weighted according to its coherency measure. Once the optimal migration velocity is used, it is assumed that in the actual specular direction, the coherency measure (semblance) along reflection events, from all available opening angles and opening azimuths, is larger than that along non-specular directions. The computed direction-dependent semblance attribute is designed to operate as an imaging filter which enhances specular migration contributions and suppresses all others in the final migration image. The ability to analyse the structural properties of the image points by the multi-parameter common image gather allows us to better handle cases of complicated wave propagation and to improve the image quality at poorly illuminated regions or near complex structures. The proposed method and some of its practical benefits are demonstrated through detailed analysis of synthetic and real data examples.

Original languageEnglish
Pages (from-to)1453-1467
Number of pages15
JournalGeophysical Prospecting
Volume62
Issue number6
DOIs
StatePublished - 1 Nov 2014

Keywords

  • Depth migration
  • Imaging
  • Noise rejection

Fingerprint

Dive into the research topics of 'Image enhancement by multi-parameter characterization of common image gathers'. Together they form a unique fingerprint.

Cite this