Velocity analysis on common offset GPR data: A deep learning approach

Giacomo Roncoroni, Matteo Dossi, Emanuele Forte, M. Pipan, Luca Bortolussi18th 2020 · DOI

Abstract

We implemented a Deep Learning algorithm to estimate the subsurface EM velocity field from common offset GPR profiles. The Deep Learning approach is based on a Bi-Directional Long Short-Term Memory (LSTM) Neural Network (NN) architecture trained on simple synthetic profiles randomly generated. The trained network is then applied to each A-Scan of 2D or even 3D GPR datasets. We trained the network on a synthetic dataset with different numbers of reflectors, wavelets, Signal-to-Noise ratios. The application of the network to synthetic and field data successfully predicts the velocity model and provides a computationally effective alternative to classic methods.
BibTeX
@inproceedings{roncoroni2020velocity,
  author = {Giacomo Roncoroni and Matteo Dossi and Emanuele Forte and M. Pipan and Luca Bortolussi},
  title = {{Velocity analysis on common offset GPR data: A deep learning approach}},
  booktitle = {18th},
  year = {2020},
  doi = {10.1190/gpr2020-101.1},
}

Loading related content...

Figure