Funding organisation: European Partnership on Metrology
Contact Person: Kostas Giannis
Summary of the overall project:
Precise calibration of microscopes at the atomic level is crucial for advances in nanotechnology, materials science, and molecular biology. Conventional calibration methods are often manual and time-consuming, and they do not offer the level of precision and reproducibility required for cutting-edge research. This thesis will investigate, standardized approach to calibrating microscopes at the atomic level using neural networks (NN) and generative adversarial networks (GANs). By automating the calibration process, this research aims to improve the accuracy, efficiency, and reliability of atomic-scale imaging, thereby fostering scientific discoveries and innovation. High-resolution electron microscopes and scanning probe microscopes are essential tools for visualizing and analyzing materials at the atomic level. However, their precision depends heavily on careful calibration. Current methods are prone to human error and inconsistencies, which limit their effectiveness and reproducibility. Crystalline standards with self-organized structures are often used for calibration. These structures must be identified and distinguished from unusable ones for further processing. This work addresses the urgent need for an automated, standardized calibration method that can be used for the post-processing of crystalline standards.
Fig.: Schematic representation of the project plan