teigen

Test Image Generator. The basic concept of the algorithm includes the definition of objects to generate, the generation of the framework of the fiber structure, the surface representation, the quantitative description, the volume representation, and finally the file storage.

user manual

Installation

On Linux, OS X or Windows the conda system can be used:

conda install -c mjirik -c simpleitk teigen

Or you can use Windows installer

Parameters

python teigen.main -p parameters.yaml

Detailed description of paremeters can be found in user manual

Paper datasets and figures

All experiment configuration

Run all experimerimens and generate all data

Figures generator script

• Dataset1 was used for generating image stacks that were measured on the micro-CT console.
• Dataset2 was used for generating image stacks for comparing known surfaces and volumes with numerically estimated surfaces and volumes.
• Dataset3 was used for generating image stacks for estimating the sensitivity of the estimates upon various numbers of objects and resolution of the measurement.

Sample output

Unconnected tubes with volume fraction 22 %. Number of elements is 335.

Tubes with overlap. Volume fraction 48 %. Number of elements is 400.

Porosities with no overlap. Volume fraction 89 %, number of elements is 400.