Full text loading...
-
Ground Roll Attenuation Using Generative Adversarial Network
- Publisher: European Association of Geoscientists & Engineers
- Source: Conference Proceedings, 81st EAGE Conference and Exhibition 2019, Jun 2019, Volume 2019, p.1 - 5
Abstract
We propose to use a generative adversarial network for ground roll attenuation. Because of the non-stationary property of seismic data and associated ground-roll noise, we create training data sets using regularised non-stationary regression approach. The basic idea is to train the network using few shot gathers such that the network can learn the weights associated with the noise attenuation for the training data sets which can then be applied to the test data sets to obtain the desired signal. This approach gives results similar to the regularised non-stationary regression approach at a significantly reduced cost. Proposed algorithm is computationally efficient and cost effective as compared to the conventional techniques. Field data example verifies the effectiveness of the proposed approach.