Enhancing X-ray micro-CT analysis for detecting voids and carbon fibre features in fibre-reinforced cementitious composites using advanced 3D Gaussian filtering
Yunyun Tao, Ziyuan Wang, S.A. Hadigheh
Abstract
With the growing applications of carbon fibre-reinforced cementitious composites (carbon-FRCC), X-ray micro-computed tomography (micro-CT) has become essential for evaluating their internal microstructures. However, analysing carbon fibres in the cementitious matrix using X-ray imaging is challenging due to poor phase contrast and image noise. This study contributes to the development and application of a 3D second-order Gaussian kernel in X-ray micro-CT analysis, specifically tailored for detecting voids and carbon fibres in cementitious materials. A novel Anisotropic Gaussian filtering was developed to achieve precise and sharp boundaries between fibres and the matrix. Higher accuracy was verified, which outperformed conventional methods across performance metrics, confirming the effectiveness of the proposed segmentation framework. The study also demonstrated the successful application of the proposed Gaussian filtering for void analysis and detecting the spatial location of carbon fibre bundles. These advancements led to more reliable feature analysis, expanding the applicability of X-ray micro-CT for examining internal features in carbon-FRCC.