Denoising electron holograms using the wavelet hidden Markov model for phase retrieval - Applications to the phase-shifting method

Abstract

Noise reduction using the wavelet hidden Markov model (WHMM) was applied to electron holograms of a LaFeO3/SrTiO3 specimen, which was subjected to phase retrieval by the phase-shifting method. A systematic study revealed the optimal conditions regarding the electron dose and number of holograms used for phase retrieval, for which the effect of denoising was most pronounced. The denoised holograms revealed the magnitude of the phase shift at the LaFeO3/SrTiO3 interface, while the original, noisy holograms did not allow for accurate measurements. The observations demonstrate the usefulness of denoising using the WHMM.