Automatic electron hologram acquisition of catalyst nanoparticles using particle detection with image processing and machine learning

Fumiaki Ichihashi,Akira Koyama,Tetsuya Akashi,Yasukazu Murakami,Shoko Miyauchi,Yoshio Takahashi,Ken'Ichi Morooka,Hajime Hojo,Toshiaki Tanigaki,Hiroyuki Shinada,Hisahiro Einaga

doi:10.1063/5.0074231

Fumiaki Ichihashi, Akira Koyama + Show 9 more

Open Access

https://doi.org/10.1063/5.0074231

Copy DOI

Abstract

To enable better statistical analysis of catalyst nanoparticles by high-resolution electron holography, we improved the particle detection accuracy of our previously developed automated hologram acquisition system by using an image classifier trained with machine learning. The detection accuracy of 83% was achieved with the small training data of just 232 images showing nanoparticles by utilizing transfer learning based on VGG16 to train the image classifier. Although the construction of training data generally requires much effort, the time needed to select the training data candidates was significantly shortened by utilizing a pattern matching technique. Experimental results showed that the high-resolution hologram acquisition efficiency was improved by factors of about 100 and 6 compared to a scan method and a pattern-matching-only method, respectively.

Highlights

Electron holography is a microscopy technique that analyzes the amplitude and phase of electron waves that have passed through a sample by utilizing information on their interference.7 If the sample is thin enough or can be assumed as a simple projection model, the electromagnetic field can be measured by analyzing the phase shift
To enable better statistical analysis of catalyst nanoparticles by high-resolution electron holography, we improved the particle detection accuracy of our previously developed automated hologram acquisition system by using an image classifier trained with machine learning
The detection accuracy of 83% was achieved with the small training data of just 232 images showing nanoparticles by utilizing transfer learning based on VGG16 to train the image classifier

Summary

Introduction

Electron holography is a microscopy technique that analyzes the amplitude and phase of electron waves that have passed through a sample by utilizing information on their interference.7 If the sample is thin enough or can be assumed as a simple projection model, the electromagnetic field can be measured by analyzing the phase shift. To enable better statistical analysis of catalyst nanoparticles by high-resolution electron holography, we improved the particle detection accuracy of our previously developed automated hologram acquisition system by using an image classifier trained with machine learning.

Results

Conclusion