XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3 1029
containing both mineral content and surface ratios for each
of the thousands of particles analysed.
It was a combination of MATLAB (MathWorks) and
Avizo (ThermoFisher scientific) software algorithms, based
on the works of Reyes et al. (2018) and Salinas-Farran et al.
(2022). The different phases in the sample (which for this
study were goethite, hematite, and gangue) were segmented
from one another. Particle size and mineral grade were
measured using standard image processing methods, which
involved counting voxels (which is the 3D equivalent to a
pixel) for each phase. Then, by considering the density of
each component, the volumetric fractions were converted
into mass ratios, obtaining the 3D mineral grade. An exam-
ple of this calculation can be found in equation (1):
Figure 1. Particle size distribution of the iron ore sample used for surface liberation
analysis
Figure 2. Stereological effect and its associated errors when describing properties of a
3D object with a 2D representation (adapted from Reyes et al., 2018)
Previous Page Next Page