Evaluation of Bayesian Classifier for Salt Dome Detection using Texture analysis

Research output: Contribution to conferencePaperpeer-review

Abstract

Seismic data analysis is an important process for oil and gas exploration. Usually, this data contains large noise and amplitude variations which hinders inferring geologic features such as salt domes accurately. The automatic detection of salt domes is always encouraged because it compensates for the limitations of manual picking, and it serves as an indicator of the existence of oil and natural gas reservoirs. In this paper, we investigate and study the performance of three texture feature extraction methods for detecting salt domes using Bayesian classifier. The Bayesian classifier is trained using one source of seismic data and evaluated using another seismic data source. Accordingly, the study tests statistically the machine learning model generalization using different seismic sources as well as the effectiveness of the texture features in this domain. Experiments confirm that the machine learning models can achieve perfect recognition accuracy using Gabor features on a single seismic source. However, these trained machine-learning models require additional training to generalize to other seismic sources.

Original languageEnglish
StatePublished - 2019

Bibliographical note

Publisher Copyright:
© 2019 Institution of Engineering and Technology. All rights reserved.

Keywords

  • DCT
  • GLCM
  • Gabor filters
  • Machine learning
  • Salt domes
  • Seismic data analysis
  • Texture Analysis

ASJC Scopus subject areas

  • Electrical and Electronic Engineering

Fingerprint

Dive into the research topics of 'Evaluation of Bayesian Classifier for Salt Dome Detection using Texture analysis'. Together they form a unique fingerprint.

Cite this