[M-IS16] Planetary Volcanology

[MIS16-04] Identification process of structure and lithology of volcanic outcrops from visual characteristics

*Nobuo Geshi¹, Junichi Haruyama², Rina Noguchi³, Keiichiro Fujimoto², Motomaro Shirao, Daigo Shoji², Kodai Ikeya⁴ (1.Geological Survey of Japan, The National Institute of Advanced Industrial Science and Technology, 2.Japan Aerospace Exploration Agency, 3.Faculty of Science, Niigata University, 4.School of Engineering, Tokai University )

Keywords:geological field survey, outcrop observation, lithology identification, volcano

The primary process of the field survey of geological outcrops is the identification of the geological structures and compositions of the outcrop based on the visual characteristics. Then, field workers can select the survey sites for more detailed information and sample collection based on the geological structure. Therefore, the identification of the geological structure and composition by visual observation is an important step in the early stage of the geological fieldwork. Automatic identification of outcrop structure and lithology from optical camera images could reveal the geological structures of a wide range of outcrops from a large volume of images that are beyond human processing capacity. This could enable efficient identification of important observation points based on this information. Such a technology could be extremely useful, especially in the exploration of polar regions, the Moon, and planets.
The goal of this study is the automatic identification of the outcrop structure and lithology based on the optic images. Outcrops in volcanic regions were targeted for application to exploration on the moon and planetary surfaces. In the first step, we analyze the working process of geologists to identify the key characteristics of the optical images of geological outcrops. We also aim to organize the thoughts and procedures of geologists during their field observation. We try to reflect the results in image processing and identification algorithms. In this presentation, we report on the results of this process of organizing the thoughts and procedures of the researcher when observing an outcrop.
First, we conducted a brief questionnaire to identify the key characteristics of the outcrop for geologists to identify the geological structure and lithology. The results showed that geologists focus on the color and surface texture of the outcrop to identify the lithology. The importance of color or surface texture depends on the target and its condition. The focus points required to identify the lithology of each outcrop are almost always consistent among geologists with a certain level of field survey experience. Based on these findings, we compared the results of automatic discrimination of lithology by image processing with actual field observation of outcrops by geologists. Based on these results, we examined which features should be identified by image processing to understand the lithology of the outcrops.
In unknown outcrops, geologists recognize the lithology by selecting locations on the outcrop surface that best match the characteristics of the known lithology. Furthermore, it was found that geologists infer the distribution of lithology by starting from locations where the lithology is recognized, and gradually extending to locations where the features are unclear. In particular, when the surface is coated with dust or secondary precipitates, it is clear that the geologist recognizes the lithology by taking into account which outcrop features are obscured by the coating and which secondary features are added.