A skeletonization based image segmentation algorithm to isolate slender regions in 3D microstructures

Abstract

The work proposes an image segmentation algorithm that isolates slender regions in three-dimensional microstructures. Characterizing slender regions in material microstructures is an extremely important aspect in material science because these regions govern the macroscopic behavior of materials for many applications like energy absorption, activation of metamaterials, stability of high temperature filters, etc. This work utilizes skeletonization method to calculate centerline of the microstructure geometry followed by a novel pruning strategy based on cross-sectional area to identify slender regions in the microstructure. 3D images of such microstructures obtained from micro-CT often suffer from low image resolution resulting in high surface noise. The skeleton of such an image has many spurious skeletal branches that do not represent the actual microstructure geometry. The proposed pruning method of cross-sectional area is insensitive to surface noise and hence is a reliable method of identifying skeletal branches that represent the slender regions in the microstructure. The proposed algorithm is implemented on a test case to showcase its effectiveness. Further it is implemented on a 3D microstructure of ceramic foam to identify the slender regions present in it. It is shown that the method can be used to segment slender regions of varying dimensions and to study their geometric properties.