Regensburg 2025 – wissenschaftliches Programm
Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
O: Fachverband Oberflächenphysik
O 16: Scanning Probe Techniques: Method Development
O 16.4: Vortrag
Montag, 17. März 2025, 15:45–16:00, H25
From experiments to insights: processing tool for SPM images with periodic pattern — •Farzin Irandoost1, Fillippo Federici Canova2, Tobias Dickbreder3, Franziska Sabath3, Angelika Kühnle3, and Adam S. Foster1,4 — 1Department of Applied Physics, Aalto University, Helsinki, Finland — 2Nanolayers Research Computing Ltd., London, England — 3Physical Chemistry I, Bielefeld University, Germany — 4Nano Life Science Institute (WPI-NanoLSI), Kanazawa University, Kanazawa, Japan
Big datasets of Scanning Probe Microscopy (SPM) images are potentially valuable, but robust algorithms are required for preprocessing them due to the high levels of defects and noise introduced during experiments. These issues often render many images unusable, especially for in-liquid SPM studies.
As part of a study on hydration patterns using a dataset of in-liquid calcite, we developed a versatile workflow to clean the data and extract features for further analysis. This workflow automatically corrects non-linear defects, ensuring the outputs closely resemble ideal periodic patterns. Consequently, many previously discarded raw images can be recovered to prepare a large, clean dataset ready for analysis. Afterward, the features of interest could be extracted using pattern decomposition facilitated by Fourier transforms.
This approach provides access to invaluable information about the lattice and hydration patterns for our study. Additionally, it offers a versatile tool for broader analyses of images with periodic structures.
Keywords: Scanning prob microscopy; Image processing; Defect removal; Periodic patterns; Feature extraction