The downscaling code allows to enhance X-ray computed tomography (XCT) images of heterogeneous porous sample and extract pore space geometry without performing a preliminary noise filtering. The code is accurate when the porous sample is chemically effectively homogeneous across the solid phase and the void phase and the uncertainties of the imaging experiment originate only from the photon shot-noise and the unresolved micro-porosity or from the partial volume effect. Due to the noise, the intensity value of the output data varies even though the sample is chemically homogeneous. The noise can be described by a Poisson probability density function (PDF), which can be approximated by a Gaussian distribution. We assume that the signal-to-noise ratio is high enough to allow identification of the solid and void phases as maximums on the data intensity histogram. The geometrical features of the unresolved micro- porosity are uncertain, but information about the porosity is still preserved in the image intensity. We define the unresolved micro-porosity as a porous phase. The main idea of the downscaling code is to map the low- resolution gray-scale image into a high-resolution binary image, while preserving the maximum information regarding the unresolved porosity, then extract the multi-scale pore distribution by segmentation of the void space of the binary image on interconnected void regions bounded by the solid phase.

The complete user's manual is located in the file Manual.pdf

Required Python libraries:

PIL, numpy, scipy, skimage, os, multiprocessing, ctypes