Image Segmentation Variants for Semi-Automated Quantitative Microstructural Analysis with ImageJ

Abstract

Abstract Porosity, pore distribution, mean grain size, and grain size distribution determine the mechanical and physical properties of ceramics. The quantitative structural analysis is therefore essential for the characterization of sintered materials. A semi-automated structural analysis requires a preceding image segmentation step in which all pixels are divided into respective objects to be examined so that they can be clearly assigned to a microstructural constituent. The present work analyzes the watershed transformation, IsoData, and WEKA algorithm image segmentation methods with regard to a grain size and pore characterization using light microscope micrographs of solid-state sintered silicon carbide (SSiC). The open source software ImageJ is used for image segmentation and detection. It does not just provide a quick quantification of the microstructural constituents but can also be extended with a considerable number of plugins, thus providing great flexibility when working on image analysis tasks.