Automatic channel detection using deep learning
Nam H. Pham, Sergey B. Fomel, Dallas B. Dunlap · 2018
We propose a method based on an encoder-decoder convolutional neural network for automatic channel detection in seismic images. We use SegNet and Bayesian SegNet architecture borrowed from computer vision. We train the network on a 3D synthetic dataset and then apply it to field data. We test the proposed approach on a 3D field dataset from the Browse Basin, offshore Australia. Applying the weights estimated from training on a 3D synthetic dataset to a 3D field dataset accurately identifies the channel geobodies without the need for any human interpretation on seismic attributes. Presentation Date: Wednesday, October 17, 2018 Start Time: 8:30:00 AM Location: 204B (Anaheim Convention Center) Presentation Type: Oral