Login / Signup

Evaluation of image distortion in SEM by using a dot-array-based certified reference material.

Kazuhiro KumagaiAkira Kurokawa
Published in: Microscopy (Oxford, England) (2021)
This paper presents a quick and straightforward method to evaluate image distortion in scanning electron microscopy using a certified reference material (CRM) as a test specimen. The CRM has a square dot-array structure, whose dot-pitch has an accredited value. By calculating the distance between each dot of the CRM via image analysis, we can detect the distortion in the image as variations of dot interval. Furthermore, by considering the uncertainty of the certified value, it is possible to quantitatively evaluate the significance of the distortion in the image. This method enables us to easily estimate the uncertainty from image distortion, which can improve the reliability of measurement by scanning electron microscopy.
Keyphrases
  • electron microscopy
  • deep learning
  • high resolution
  • energy transfer
  • high throughput
  • magnetic resonance imaging
  • computed tomography
  • machine learning
  • quantum dots