A casting method using contrast-enhanced diethylphthalate for micro-computed tomography of snow

Abstract

Casting snow is necessary to prevent metamorphism and deformation prior to X-ray micro-computed tomography (μCT) imaging. Current methods are insufficient for large-scale field sampling of snow due to safety considerations associated with the casting medium and/or lengthy sample preparation times. Here, a casting method using contrast-enhanced diethylphthalate (DEP) for μCT of snow is presented. The X-ray contrast of DEP is enhanced with barium titanate nanoparticles (BaTiO3) and iodine (I2). A partially unsupervised, three-phase segmentation method utilizing traditional Gaussian smoothing followed by a three-step process to address transition voxels is also presented. Synthetic images derived from real snow samples are used to evaluate the segmentation method with various configurations of trapped air bubbles. Real snow samples spanning a range of specific surface areas (SSAs) (8–28 m2 kg−1) and densities (135–463 kg m−3) are used to assess the performance of the segmentation method on real, cast samples. The method yields SSA, density and correlation length errors of less than 10% for synthetic images with air bubble surface areas less than 333 m−1 per sample volume for eight of the nine snow samples. For eight of the nine cast samples, the method yields errors of less than 10% for all three parameters.

Ice-Core Micro-CT Image Segmentation With Deep Learning and Gaussian Mixture Model

“…Then the microstructure parameters such as specific surface area (SSA) were computed from the segmented images. It was shown that both of the methods were prone to hardware setup, but the energy-based (graph cut) method is not subjective to the choice of a human user and it benefits from local spatial information, with less affection of beam hardening [32]- [34].…”