Affiliation:
1. University of Cambridge
2. University of Nottingham
Abstract
The effects of time-varying measurement noise on transmission matrix
acquisition processes are considered for the first time, to our
knowledge. Dominant noise sources are discussed, and the noise
properties of a typical interferometer system used for characterizing
a multimode fiber transmission matrix are quantified. It is
demonstrated that an appropriate choice of measurement basis allows a
more accurate transmission matrix to be more quickly obtained in the
presence of measurement noise. Finally, it is shown that
characterizing the noise figure of the experimental system allows the
inverse transmission matrix to be constructed with an ideal amount of
regularization, which can in turn be used for optimal image
acquisition.
Funder
UK Research and Innovation
Department of Chemical Engineering and
Biotechnology, University of Cambridge
Subject
Computer Vision and Pattern Recognition,Atomic and Molecular Physics, and Optics,Electronic, Optical and Magnetic Materials