SEISMIC INVERSION
Physics-based
Machine Learning
Elastic
FWI
Conventional Reflection SI
Convolution Neural Network (CNN)
Physics Informed Neural Network (PINN)
Anisotropic velocity models
VORT
TORT
VTI
TTI
Conventional
Reflection SI Workflow
Wave Path Tomo
Wave Path Tomo
Standard Q-RTM
Image-domain LS-Q-RTM
Data-domain LS-Q-RTM
Elastic
FWI
FWI + LS RTM
85 Hz Simultaneous Inversion
85 Hz Kirchhoff PSDM
85 Hz Simultaneous Inversion (reflectivity)
85 Hz Kirchhoff PSDM
Wave Path Tomo
Data-domain LS-Q-RTM
Wave Path Tomo
Wave Path Tomo
Field data
Q-RTM
Q-RTM
Conventional imaging workflow
Simultaneous
Conventional imaging
inversion
(FWI + LSRTM)
CNN
PINN
PHYSICS
(NN - FD)
(PINN - FD)
(PINN)
Finite Difference Modelling (FD)
Hover your cursor to see wave propagation
Training data
True model
Initial model
Predicted model
True model
Initial model
PINN-based WRI
FWI Inverted from Initial model
FWI Inverted from PINN-based WRI