Stochastic Geometry Models

We generate realistic structures, controlled by just a few parameters. We analyze image data to ensure similarity to the real structure.

Model realizations are useful to

• Simulate material properties
  
• optimize structures
  
• explore effects of structural variations without actually creating the material
  
• determine RVE sizes
  
• generate synthetic image data with ground truth
  

The basis are models from stochastic geometry such as Boolean models, straight line processes and mosaics. Macroscopic homogeneity is achieved by using only stationary models. This means that it does not matter at which point the sample was taken.