Premium
Evaluation of an Advanced Benchtop Micro‐Computed Tomography System for Quantifying Porosities and Pore‐Size Distributions of Two Brazilian Oxisols
Author(s) -
Vaz Carlos M. P.,
Maria Isabella C.,
Lasso Paulo O.,
Tuller Markus
Publication year - 2011
Publication title -
soil science society of america journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.836
H-Index - 168
eISSN - 1435-0661
pISSN - 0361-5995
DOI - 10.2136/sssaj2010.0245
Subject(s) - resolution (logic) , tomography , image resolution , materials science , segmentation , porosity , sample (material) , region of interest , artificial intelligence , biomedical engineering , computer vision , mathematics , computer science , optics , physics , composite material , medicine , thermodynamics
Recent advances in benchtop Micro‐Computed Tomography (Micro‐CT) provided the motivation to thoroughly evaluate and optimize scanning, image reconstruction/segmentation and pore space analysis capabilities of a SkyScan 1172 Micro‐CT system and associated SkyScan Analyzer software package (SkyScan, Belgium). To demonstrate applicability to soil research the project was focused on determination of porosities and pore‐size distributions of two Brazilian Oxisols from segmented computed tomography (CT)‐data. Effects of metal filters and various acquisition parameters (e.g., total rotation, rotation step, and radiograph frame averaging) on image quality and acquisition time were evaluated. Impacts of sample size and scanning resolution on CT‐derived porosities and pore‐size distributions were illustrated. Best image quality was achieved when an aluminum/copper filter (0.5/0.04 mm thickness) was positioned between sample and x‐ray detector, the sample was rotated 360° in 0.3° steps and the number of averaged projections per rotation step was larger than 15. As with all other CT systems there was a tradeoff between maximum observable sample volume and achievable resolution. Even for soil samples scanned at the highest considered resolution of 3.7 μm (1.0 μm is the maximum achievable resolution) it was apparent that CT‐derived porosities vastly underestimated physically measured apparent porosity values. In addition, it was shown that dense mineral particles can be discriminated from slightly attenuating elements and minerals. While image acquisition and reconstruction yielded excellent grayscale CT‐data, the supplied software lacked advanced image segmentation algorithms and morphometric pore space analysis capabilities.