Stereological correction method based on sectional texture analysis for the liberation distribution of binary particle systems

Abstract

Accurate assessment of liberation state of ore sample is important in mineral processing. In practice, the ore sample is mounted in resin, sectioned, and polished before its liberation state is measured by sectional analysis. This approach typically overestimates the degree of liberation in two-dimensional (2D) measurements. Several models have been proposed to overcome this stereological bias and correct this error in well-examined samples. However, their versatility remains poorly understood. Herein, a stereological correction method was developed for the liberation distribution. First, the complexity of particle sectional texture was assessed using the fractal dimension of the image intensity. Next, the magnitude of stereological bias and 2D measureable parameters such as fractal dimension was correlated by an all-encompassing simulation of various texture types. Finally, stereological correction indices were exclusively estimated from 2D measureable parameters obtained through the correlation. This model exhibited high versatility based on the all-encompassing simulation. The model was validated using nine different irregularly shaped binary particle systems and assessed using areal difference and maximum difference between liberation distribution curves in two and three dimensions. This error indices improved by approximately 80% for the former error index and 90% for the latter index.