Exploring Blob Detection to Determine Atomic Column Positions and Intensities in Time-Resolved TEM Images with...

Area of Science:

• Materials Science
• Electron Microscopy
• Image Analysis

Background:

• In situ transmission electron microscopy (TEM) with direct electron detection enables millisecond temporal resolution for atom-scale dynamics.
• High temporal resolution in TEM imaging leads to poor signal-to-noise ratios, complicating the analysis of atomic column positions and intensities.
• Characterizing fluxional behavior in materials at the atomic scale is challenging due to image noise.

Purpose of the Study:

• To develop a robust image processing approach for analyzing high-noise, high-temporal-resolution TEM image series.
• To improve the precision of determining atomic column position and intensity in dynamic nanomaterials.
• To adapt computer vision techniques for advanced materials characterization.

Main Methods:

• Implementation of a noise-robust blob detection algorithm tailored for TEM image series.
• Application of the blob detection method to noisy images of nanoparticle systems.
• Comparison of the blob detection approach with existing algorithms for image analysis.

Main Results:

• The proposed blob detection method effectively handles high noise content in TEM images.
• The algorithm outperforms other methods in determining atomic column position and intensity.
• Enhanced precision in analyzing atom-scale dynamics of materials is achieved.

Conclusions:

• Blob detection offers a superior solution for processing noisy TEM data, particularly for dynamic systems.
• This approach facilitates more accurate characterization of atomic-scale material behavior.
• The method advances the capabilities of in situ TEM for studying materials dynamics.