3D size and shape characterization of natural sand particles using 2D image analysis

Abstract

The geometrical size and shape of sand particles are critically important for evaluating the characteristics of granular materials. These characteristics are commonly quantified via the analysis of 2D and 3D images. In the traditional procedure, a 2D particle image is captured by a digital camera only once and only limited amount information about a particle is acquired. In the present study, a framework that integrates image analysis and 3D laser scanning was developed and applied to 20 natural sand particles to study multiple acquisitions and evaluate the size and shape. The potential of single, two, three and multiple projections of sand particles to characterize morphological properties were studied and the relationships between the 3D and 2D parameters were established. The results show that the 3D morphological properties including the size (length, width, thickness, volume, and surface area) and shape (elongation, flatness, and sphericity) can be estimated using 2D morphological properties. The parameters measured on the maximum and minimum projections (from 20 acquisitions) or three orthogonal projections were found to be useful for estimating the 3D geometrical properties. A practical approach to obtain representative sand particle information from 2D image analysis is provided. The newly developed equations were verified and compared with previously published equations.