A martian case study of segmenting images automatically for granulometry and sedimentology, Part 1: Algorithm
In planetary exploration, delineating individual grains in images via segmentation is a key path to sedimentological comparisons with the extensive terrestrial literature. Samples that contain a substantial fine grain component, common at Meridiani and Gusev at Mars, would involve prohibitive effort if attempted manually. Unavailability of physical samples also precludes standard terrestrial methods such as sieving. Furthermore, planetary scientists have been thwarted by the dearth of segmentation algorithms customized for planetary applications, including Mars, and often rely on sub-optimal solutions adapted from medical software. We address this with an original algorithm optimized to segment whole images from the Microscopic Imager of the Mars Exploration Rovers. While our code operates with minimal human guidance, its default parameters can be modified easily for different geologic settings and imagers on Earth and other planets, such as the Curiosity Rover's Mars Hand Lens Instrument. We assess the algorithm's robustness in a companion work. © 2013 Elsevier Inc.
Publication Source (Journal or Book title)
Karunatillake, S., McLennan, S., Herkenhoff, K., Husch, J., Hardgrove, C., & Skok, J. (2014). A martian case study of segmenting images automatically for granulometry and sedimentology, Part 1: Algorithm. Icarus, 229, 400-407. https://doi.org/10.1016/j.icarus.2013.10.001