2015/08/16 by Daniel Fra̧czek, D. Frączek, W. Olchawa +6 · 1 citation
Biochemistry, Genetics and Molecular Biology · Computer Science · Physics and Astronomy · #Cell Image Analysis Techniques #FOS: Physical sciences #Image Retrieval and Classification Techniques #Medical Image Segmentation Techniques #Statistical Mechanics (cond-mat.stat-mech) #cond-mat.stat-mech
paper · pdf · doi:10.48550/arxiv.1508.03857
10 pages, 6 figures, clarified version
openalex publication_date 2015/08/16 · arxiv created 2015/11/16 · arxiv updated 2015/11/18 · openalex created_date 2016/06/24 · openalex updated_date 2026/07/28
The simple entropic method to statistical reconstructing of heterogeneous three-dimensional media from a single two-dimensional image is briefly reported. We apply the entropic descriptor quantifying spatial inhomogeneity that depends on length scale. The multiscale reconstructing procedure is boosted when we start from the synthetic 3D configuration. It is randomly generated with the overlapping spheres of a radius depending on the structure considered. In our scenario, the simulated annealing terminates when all assigned temperature loops are completed. Thus, the quality of the results for different samples can be fairly compared. The reliable 3D reconstructions of porous sandstone, ceramics and carbonate samples are obtained. They suggest the entropic method is a promising approach, which offers a kind of compromise between the computational efficiency and the accuracy of the statistical reconstructions. Furthermore, this approach is versatile enough to be extended for any multiphase medium.