The segmentation of Focus Ion Beam - Secondary Electron Microscope (FIB-SEM) data of highly porous media is an unsolved problem so far. The imaging method causes several artifacts preventing the efficient and objective identification of the solid/void boundary. Therefore, different algorithms to segment the matrix from the void phase have been developed. However, they have several limitations regarding the type of the structure and the porosity level.
One of these segmentation algorithms is described by Torben Prill in the paper “Morphological Segmentation of FIB-SEM Data of Highly Porous Media” and it was validated using a FIB-SEM simulated data obtaining overall error below standard methods.
Extensions of this work are possible by including new synthetic data with more structures and more than one solid phase, and through the development of more detectors in the tool for simulating FIB-SEM data. The resulting synthetic data sets will be used to validate new preprocessing and segmentation algorithms.