Development of a computer vision-based measuring system for investigating the porous media structure

Porous materials are widely used in a variety of scientific and technological fields; therefore, a large number of publications deal with the challenges related to studying the porous media microstructure. Porosity investigation methods based on the measurement of certain physical properties of the medium are usually very complex, expensive, cumbersome, require harsh experimental conditions and are unable to determine pore size distribution. On the other hand, modern optical measuring techniques are also not universal and often need significant adaptation. In this regard, a strategy is developed that formalizes the process of direct porosity measurement of the investigated specimen using a series of its images. The proposed method performs image denoising, aggregates separation, measurement of microstructure elements and mathematical analysis of acquired data. Numerical experiments performed using CT images of a porous ceramic sample suggest the reliability of the developed technique.