Segmenting images automatically for granulometry and sedimentology: a martian case study

In a companion work, we bridge the gap between mature segmentation software used in terrestrial sedimentology
and emergent planetary segmentation with an original algorithm optimized to segment
whole images from the Microscopic Imager (MI) of the Mars Exploration Rovers (MER). In this work,
we compare its semi-automated outcome with manual photoanalyses using unconsolidated sediment
at Gusev and Meridiani Planum sites for geologic context. On average, our code and manual segmentation
converge to within ~10% in the number and total area of identified grains in a pseudo-random, single
blind comparison of 50 samples. Unlike manual segmentation, it also locates finer grains in an image with
internal consistency, enabling robust comparisons across geologic contexts. When implemented in Mathematica-
8, the algorithm segments an entire MI image within minutes, surpassing the extent and speed
possible with manual segmentation by about a factor of ten. These results indicate that our algorithm
enables not only new sedimentological insight from the MER MI data, but also detailed sedimentology
with the Mars Science Laboratory’s Mars Hand Lens Instrument.