3D Collection of Binary Images

Project Cover Image


This project presents several 3D binary geometries (artificially created) that represent a wide variety of systems. We represent the pore-space with zeros and the solid voxels with ones. We have included sphere packing (eroded and dilated to create pore structure similar in shape but a variation of porosity), single fractures, packings with overlapping spheres, shale reconstructions and fractures with permeable walls, among others. Most of the samples are presented as a cube of side length 256 voxels, some include an additional one with 480 voxels. These are dimensionless and can be scaled to any physical length of interest. Each image was checked for connectivity in z-direction (third axis) and the isolated components were removed. Each dataset includes a binary file (in hdf5 format), a 3D picture showing the voxelized geometry, and a plot showing cross-sectional planes of the domain. These images were plotted using Vedo and PyVista, respectively. --------------------------------------- To load the geometry in python: --------------------------------------- from hdf5storage import loadmat bin = loadmat('374_01_00_256.mat')['bin'] --------------------------------------- To load the geometry in matlab: --------------------------------------- load('374_01_00_256.mat')


Usage Information



  • Ying Yin (UT)
  • Masa Prodanovic (The University of Texas at Austin)
  • Alex Gigliotti (The University of Texas at Austin)
  • Nicholas Lubbers (LANL)
  • Hasan J Khan (King Fahd University of Petroleum & Minerals)


July 6, 2021


ODC-BY 1.0

Digital Object Identifier


Data Citation

Download Project

The downloadable archive contains all project data; the size of the archive file for this project is 0.52 GB.