[25p-P02-1~2] 6.5 Surface Physics, Vacuum

[25p-P02-2] Automatic Extraction of Atomic Positions in Field Ion Microscope Images by Machine Learning

〇(M1)Tadasuke Okazawa¹, Yamada Mizuki¹, Yuga Azechi², Hideto Nakai², Shigekazu Nagai¹, Kouichi Hata¹ (1.Graduate School of Eng., Mie Univ., 2.Faculty of Eng., Mie Univ.)

Keywords:Field Ion Microscopy, machine learning, field evaporation

The atoms that are imaged in the FIM image are only those that protrude on the surface due to the field strength required for the ionization of the imaging gas, and not all surface atoms of the sample can be observed. Therefore, there have been attempts to observe all the surface atoms by recording FIM images while removing them by field evaporation. In this study, we attempted to automate and speed up the above process by using machine learning to extract the surface atom positions.