Full text loading...
-
U_Net & Residual Neural Networks for Seismic Fault Interpretation
- Publisher: European Association of Geoscientists & Engineers
- Source: Conference Proceedings, 81st EAGE Conference and Exhibition 2019, Jun 2019, Volume 2019, p.1 - 5
Abstract
Deep learning has showed success in various industries for different processing tasks. However, it is still in the initial stage of the application of seismic exploration. In this paper, we propose to apply deep learning technology to detect seismic faults. The residual neural network (Resnet) is used to build the U-net model for seismic fault detection. And multi-scale and multi-level feature are extracted from seismic data. In the training stage, we use field dataset to train the U_net model. Finally, field data experiment demonstrates that our proposed U-net model is feasible for seismic fault detection. In the meantime, our research will help promote the intelligent development of seismic faults interpretation and provides a reference for further studies of seismic data processing and interpretation.