On simulated blurred ISS imagery, U-Net-only restoration reduced pose-estimation angular error by about 72% relative to no preprocessing.
Point-Spread Function of the Optics in Scanning Electron Microscopes
1 Pith paper cite this work. Polarity classification is still indexing.
abstract
Point-spread function of the probe forming optics ($PSF_{optics} $) is reported for the first time in an uncorrected (without multipole correctors) scanning electron microscope (SEM). In an SEM, the electron probe information is lost as the beam interacts with the specimen. We show how the probe phase information can be recovered from reconstructed probe intensity estimates. Controlled defocus was used to capture a focal-series of SEM images of $28.5\;nm $ gold ($\mathrm{Au} $) nanoparticles ($\mathrm{NPs} $) on a carbon ($\mathrm C $) film. These images were used to reconstruct their respective probe intensities to create a focal-series of probe intensities, which were the input to the phase retrieval pipeline. Using the complete description (intensity and phase) of the electron probe wavefunction at the specimen plane, we report the $PSF_{optics} $ for multiple data sets for beam energy $E\;=20\;keV\; $. This work opens up new possibilities for an alternative way of aberration correction and aberration-free imaging in scanning electron microscopy.
citation-role summary
citation-polarity summary
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
background 1representative citing papers
citing papers explorer
-
Deep Learning-Based Image Recovery and Pose Estimation for Resident Space Objects
On simulated blurred ISS imagery, U-Net-only restoration reduced pose-estimation angular error by about 72% relative to no preprocessing.