Skip to main navigation Skip to search Skip to main content

Seismic data compression using signal alignment and PCA

  • King Fahd University of Petroleum and Minerals

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

5 Scopus citations

Abstract

Principal Component Analysis (PCA) offers an optimal dimensionality reduction while maintaining the variances. A set of seismic traces data recorded by a sensor can be compressed by projecting the data to the Principal Components (PCs). The reconstruction error can be determined by choosing number of PCs. If the traces are aligned according to some references, number of PCs becomes fewer for the same preserved eigenvalues. Since the fewer PCs are required, compression ratio becomes higher and transmission cost from each sensor becomes smaller. Maximum amplitude and crosscorrelation techniques are evaluated to perform traces alignment. In the experiments, the aligned PCA achieves 12:1 compression ratio outperforming conventional PCA with 9.9:1 preserving approximately 99% of energy with reconstruction error 0.8% and 0.68%, respectively.

Original languageEnglish
Title of host publication2017 9th IEEE-GCC Conference and Exhibition, GCCCE 2017
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Print)9781538627563
DOIs
StatePublished - 27 Aug 2018
Externally publishedYes
Event9th IEEE-GCC Conference and Exhibition, GCCCE 2017 - Manama, Bahrain
Duration: 8 May 201711 May 2017

Publication series

Name2017 9th IEEE-GCC Conference and Exhibition, GCCCE 2017

Conference

Conference9th IEEE-GCC Conference and Exhibition, GCCCE 2017
Country/TerritoryBahrain
CityManama
Period8/05/1711/05/17

Keywords

  • Alignment
  • Compression
  • Cross-correlation
  • PCA
  • Seismic traces

Fingerprint

Dive into the research topics of 'Seismic data compression using signal alignment and PCA'. Together they form a unique fingerprint.

Cite this