ABSTRACT

At present, the stacking velocity mainly are acquired by manual picking on velocity spectra, which is a tedious and time-consuming process due to the growing number of the seismic data, especially 3D seismic data. To improve the efficiency in velocity picking, we have developed a new automatic velocity picking method based on deep learning. We combine two different structures of artificial neural networks, the You Only Look Once (YOLO) and Long Short-Term Memory (LSTM) for constructing an algorithm of automatic velocity picking, which are designed according to two aspects: (1) The process of velocity picking may be considered as the detection of maximum coherency values associated with primary reflections in the velocity spectra. (2) The time-velocity pairs picked from velocity spectra is a time sequence. The test with real velocity spectra from a marine seismic data set demonstrates that the deep neural network of YOLOLSTM model for velocity auto-picking is much more efficient than manual picking.

Presentation Date: Wednesday, September 18, 2019

Session Start Time: 1:50 PM

Presentation Time: 2:15 PM

Location: Poster Station 2

Presentation Type: Poster

This content is only available via PDF.
You can access this article if you purchase or spend a download.