Segmentation and Analysis of Particles in 3D

MAVI or ToolIP+MAVIkit can be used to segment particles in three-dimensional image data and determine the size and shape distributions of these objects. The length, width and height of complex-shaped particles can be robustly obtained from the edge lengths of the minimum enveloping cuboid.

Particularly pointed particles or systems of particles of very different sizes cannot always be separated directly without errors. A random forest that recognizes and joins splinters helps here.

Schematic Diagram of Graphic Particle Systems
© Fraunhofer ITWM
Particle separation process: Particles that are virtually connected are separated using image analysis. Errors are then corrected by a random forest based on geometric parameters.
Characterization of particles in 3D: Maximum Feret diameter
© Fraunhofer ITWM
Characterization of particles in 3D: Maximum Feret diameter
Characterization of particles in 3D: Maximum elongation
© ITWM
Characterization of particles in 3D: Maximum elongation
Characterization of particles in 3D: Maximal local thickness
© Fraunhofer ITWM
Characterization of particles in 3D: Maximal local thickness

Publications

  • Burgmann, S.; Godehardt, M.; Schladitz, K.; Breit, W.:
    Separation of sand and gravel particles in 3D images using the adaptive h-extrema transform. Powder Technology: 404, 117468, 2022.
  • Burgmann, S.; Godehardt, M.; Schladitz, K.; Breit, W.:
    Influence of voxel size for µCT imaging of particles on measurement accuracy.
    Construction and Building Materials: 289 (3), 123148, 2021.
  • Vecchio, I.; Schladitz, K.; Godehardt, M.; Heneka, M.:
    3D Geometric Characterization of Particles Applied to Technical Cleanliness.
    Image Analysis & Stereology: 31 (3), 2012.
  • Weber, O.; Rack, A.; Redenbach, C.; Schulz, M.; Wirjadi, O.:
    Micropowder Injection Molding: Investigation of Powder-Binder Separation Using Synchrotron-Based Microtomography and 3D Image Analysis.
    Journal of Materials Science: 46 (10), 2011.