Seminars Archive

PORE3D: a program for quantitative image analysis of porous media

Abstract
Quantitative information about the morphology of composites and their dispersion characteristics can lead to a better understanding of the structure and can be used to establish interrelationships between structure, processing effects, material properties and engineering performance. By means of new techniques such as X-ray microtomography we can probe the microstructure of samples and obtain high resolution images down to a few micrometers. Analyzing these images rich of information allows us extract useful characteristics and achieve the pore network structure. PORE3D is a program designed for processing two or three dimensional digitized gray-scale image, generated by X-ray microtomography. The goal of this code is to produce geometric analysis of the phases in the composite. This is achieved by obtaining some geometrical characterization taken from the image. The program is a research code written in MATLAB and contains the following major applications: filtering, segmentation, separation, feature extraction, visualization. In PORE3D we can load slices of 2D images, or 3D images, or 3D raw data. User can select a region of interest (ROI). There are some helpful filters (Gaussian, Wiener, etc.) for removing the noise and enhancing the contrast which can be useful for preprocessing stage. The important job to be done now is segmentation that is defining which voxels in the data set belong to which phase. There are 3 segmentation methods based on intensity thresholding and one based on k-mean clustering. Then, the separation method is used to separate all of the voxels in each phase into 3D-connected objects. Finally, user can choose his interesting characteristic among the list of features (porosity, volume, surface, etc.) which will be calculated on the selected phase and the result will be demonstrated after. In visualization section, there are some choices for the user which make observation easier both in 2D and 3D (slice movie, rendering, etc.). This code has been applied to the image analysis of materials such as rocks, chocolate, bones, cements, foams, etc. studied at the SYRMEP beam line of Sincrotrone Trieste and showed satisfactory results.