Analysis and Modeling of 3D Microstructures
Dominique Jeulin · 2013
Image analysis gives rise to 3D implementation of classical tools for morphological segmentation, where different information may be used: 3D local orientation to extract lamellar structures, multiscale filtering for segmentation of granular media, and so on. The classical tools of 2D morphological processing can often be adapted to 3D. However, because of the more demanding calculations and of the increased size of data files, it is necessary to use particularly efficient algorithms. Also, some specific 3D morphological descriptors can enrich the characterization of microstructures. This chapter illustrates these points using examples from studies of porous media, granular and metallic alloys. It develops a fast and efficient method to estimate the percolation threshold of complex microstructures. The chapter reports some results on Boolean models for primary grains composed of spheres and sphero-cylinders. Controlled Vocabulary Terms crystal microstructure; image processing; lattice theory; materials science computing; mathematical morphology