MOSAIC: an investigation of quantitative morphometrics of vesicular samples using OpenCV blob detection

The study of shape and form is essential to geology and is known as morphometrics. Morphometrics when applied to the study of volcanic rock samples helps to unlock the quantitative information hidden within them. The bubble space within the volcanic samples, known as vesicles, and their related measurements such as vesicle size distribution and vesicularity help geologists understand volcanic processes such as eruption dynamics. These quantitative metrics are difficult to evaluate effectively using existing tools and methods. A new method has been developed to analyze these metrics. The Modeling Object Structure & Analysis Information Calculator (MOSAIC) is a software application designed to produce quantitative morphometrics for geological digital images. It focuses on bubble space within the rock samples in digital images using the open-source computer vision image processing library OpenCV. This is also used in such applications as facial recognition .

MOSAIC applies a shape recognition algorithm known as blob detection to evaluate bubble geometry patterns within the digital image of the volcanic sample. MOSAIC reports quantitative morphometrics of bubble form including circularity, convexity, compactness, roughness, roundness and sphericity as well as area, volume, symmetry, and vesicularity quantities. 3D computed tomography data from a highly vesicular pumice (New Zealand) and a moderately vesicular Pele s hair sample (Hawaii) and simulated geologic 2D data were also used for comparisons between various vesicle analysis methods. This study evaluates the ability of MOSAIC to meet or exceed the ability of other commonly used morphometric analysis methods.