Unfortunately this page does not have a mobile or narrow screen view. Please switch to a desktop computer or increase the size of your browser. For tablets try flipping the screen.

Metadata Files

Data Publication

New 2D-to-3D Reconstruction of Heterogenous Porous Media via Deep Generative Adversarial Networks (GANs)

Hannah Vogel | Hamed Amiri

Utrecht University

(2024)

Descriptions

We imaged samples of Berea sandstone from Ohio (USA) using two 2D imaging techniques, backscattered electron (BSE) and optical microscopy, as well as one 3D imaging technique, X-ray (micro-)computed tomography (XCT). The goal is to employ a deep-learning-based generative model called a generative adversarial network (GAN) to reconstruct statistically equivalent microstructures in 3D from exclusively 2D training images. To evaluate the reconstruction accuracy, we conduct a visual and statistical analysis comparing reconstructions with a 3D X-ray tomography of the same sample. Our method uniquely uses true 2D images from three orthogonally oriented planes for training the model. The data are organized into 13 folders: three contain the original segmented (binary) images of Berea sandstone samples, and the other 10 folders contain data and individual figures used to create figures in the main publication. Link to GitHub containing codes: https://github.com/hamediut/2D-to3D-recon

Keywords


Originally assigned keywords
Analytical and microscopy data
2D-to-3D reconstructions
image reconstruction
microstructural quantification
generative adversarial network (GAN)
heterogenous microstructures
porous media
statistically-equivalent microstructures
backscattered electron microscopy
scanning electron microscopy
X-ray tomography
optical microscopy
Micro-CT
deep learning (DL)
artificial intelligence (AI)
Berea sandstone
Link to GitHub containing codes: https://github.com/hamediut/2D-to3D-recon
sandstone
metamorphic_rock
serpentinite
Scanning Electrone Microscope
Optical Microscope
Permeability

Corresponding MSL vocabulary keywords
backscatter electron (BSE) imaging
X-ray tomography
computed tomography (CT)
optical microscopy
Berea sandstone
sandstone
serpentinite
scanning electron microscope (SEM)
permeability

MSL enriched keywords
Technique
imaging (2D)
backscatter electron (BSE) imaging
Apparatus
X-ray tomography
imaging (3D)
computed tomography (CT)
optical microscopy
sedimentary rock
sandstone
wacke
Berea sandstone
metamorphic rock
serpentinite
electron microscopy
scanning electron microscope (SEM)
Measured property
permeability

MSL enriched sub domains i

microscopy and tomography
rock and melt physics

Source

https://public.yoda.uu.nl/geo/UU01/2L689L.html

Source publisher

Utrecht University


DOI

10.24416/uu01-2l689l


Creators

Hannah Vogel

Utrecht University

ORCID:

0009-0004-5393-0866

Hamed Amiri

Utrecht University

ORCID:

0000-0002-2981-1398


Contributors

Vogel, Hannah

Researcher

Utrecht University

ORCID:

0009-0004-5393-0866

Amiri, Hamed

Researcher

Utrecht University

ORCID:

0000-0002-2981-1398

Plümper, Oliver

Supervisor

Utrecht University

ORCID:

0000-0001-9726-0885


References

10.1029/2024JH000178

https://github.com/hamediut/2D-to3D-recon

https://doi.org/10.24416/UU01-DO6LT4


Citation

Vogel, H., & Amiri, H. (2024). New 2D-to-3D Reconstruction of Heterogenous Porous Media via Deep Generative Adversarial Networks (GANs) (Version 1.1) [Data set]. Utrecht University. https://doi.org/10.24416/UU01-2L689L


Dates

Updated:

2024-07-16T10:31:18

Collected:

2022-03-01/2024-05-14


Language

en


Funding References

Funder name: European Research Council

Funder name: NWO


Rights

Open - freely retrievable

Creative Commons Attribution 4.0 International Public License


Datacite version

1.1