Advanced Powder Technology, Vol.18, No.2, 187-213, 2007
Image processing and geometric parameters extracted from sliced image of porous biomaterial
To perform comprehensive modeling for intraflow field and transport processes in a biomaterial, a detailed probing technique is needed for profiling its interior. In our 2004 Advanced Powder Technology paper, 10 commonly adopted techniques were compared for their capability to probe the interior structure of sludge floc. However, morphological information extracted from sliced images depends on how the image was thresholded. As a continuation work of our 2004 Advanced Powder Technology paper, this work surveyed several frequently encountered problems when applying the histogram-based and bolb feature-based methods for the bilevel thresholding of the sliced images of flocs. A combined scheme using Otsu's method to obtain the porosity and the maximum convex perimeter method to probe the shapes and spatial distribution of the pores was discussed. Some geometric parameters were adopted to describe the interior structure of flocs. The probable misuse of the Sierpinski fractal dimension, due to the numerous pores and solid grains being simultaneously removed from floc image when its resolution was subsequently reduced, was highlighted.