Denoising high-amplitude cross-flow noise using curvelet-based stable principle component pursuit

TitleDenoising high-amplitude cross-flow noise using curvelet-based stable principle component pursuit
Publication TypeConference
Year of Publication2017
AuthorsRajiv Kumar, Nick Moldoveanu, Felix J. Herrmann
Conference NameEAGE Annual Conference Proceedings
Month06
Keywordscoil data, cross-flow noise, curvelet, denoising, EAGE, SPCP
Abstract

Removal of high-amplitude cross-flow noise in marine towed-streamer acquisition is of great interest because cross-flow noise hinders the success of subsequent processing (e.g. EPSI) and migration. However, the removal of cross-flow noise is a challenging process because cross-flow noise dominates steep angles and low-frequency components of the signal. As a result, applying a simple high-pass filter can result in a loss of coherent diving waves and reflected energy. We propose a stable curvelet-based principle-component pursuit approach that does not suffer from this shortcoming because it uses angle- and scale-adaptivity of the curvelet transform in combination with the low-rank property of cross-flow noise. As long as the cross-flow noise exhibits low-rank in the curvelet domain, our method successfully separates this signal component from the diving waves and seismic reflectivity, which is well-know to be sparse in the curvelet domain. Experimental results on a common-shot gather extracted from a coil shooting survey in the Gulf of Mexico shows the potential of our approach.

Notes

(EAGE, Paris)

URLhttps://www.slim.eos.ubc.ca/Publications/Public/Conferences/EAGE/2017/kumar2017EAGEdha/kumar2017EAGEdha.html
DOI10.3997/2214-4609.201701055
Presentation

https://www.slim.eos.ubc.ca/Publications/Public/Conferences/EAGE/2017/ku...

Citation Keykumar2017EAGEdha