Full waveform inversion with unbalanced optimal transport distance and a mixed Wasserstein distance

Preprint paper:

Application of an unbalanced optimal transport distance and a mixed L1/Wasserstein distance to full waveform inversion

Installation

git clone https://github.com/zzar43/FWI_OT

How to use

A Marmousi test data is saved in the folder ``marmousi_model''.

Forward modelling with the true model (parallel computing with 2 workers):

julia -p 2 make_data.jl

The received signal generated with true model are saved in the folder ``temp_data''.

Perform full waveform inversion with the L2 objective function:

julia -p 2 inversion_l2.jl

Perform full waveform inversion with the mixed Wasserstein distance objective function:

julia -p 2 inversion_mixed_ot.jl

Perform full waveform inversion with the unbalanced optimal transport distance objective function:

julia -p 2 inversion_uot.jl

The inverse results for each iterations are saved in the folder ``temp_data''.

Demo

Here is a demo: