We propose a variant of the U-Net neural network, dubbed U-SaltNet, for the task of salt body identification. We have incorporated several existing state-of-the-art modules into our neural network architecture. These modules include scSE, FPA and AG modules, etc. We validate our scheme on real data examples from the Gulf of Mexico, comparing with the results for experienced interpreters. Our results indicate the proposed neural network improves overall results and it can detect subtle salt features within promising precisions.
Presentation Date: Monday, October 12, 2020
Session Start Time: 1:50 PM
Presentation Time: 1:50 PM
Location: 351F
Presentation Type: Oral
This content is only available via PDF.
2020. Society of Exploration Geophysicists
You can access this article if you purchase or spend a download.
