Object reconstruction from multiplexed quantum ghost images using reduction technique

TL;DR

This paper introduces a measurement reduction technique for optimal object image reconstruction from multiplexed ghost images, leveraging correlations and sparsity, effective even with low photon counts and noisy conditions.

Contribution

It presents a novel reduction-based method for reconstructing ghost images considering correlations and sparsity, with analysis of different parametric processes and noise robustness.

Findings

Reconstruction is effective with low photon numbers.

The method improves noise immunity in ghost imaging.

Imaging conditions depend on parametric process type.

Abstract

We apply the measurement reduction technique to optimally reconstruct an object image from multiplexed ghost images (GI) while taking into account both GI correlations and object image sparsity. We show that one can reconstruct an image in that way even if the object is illuminated by a small photon number. We consider frequency GI multiplexing using coupled parametric processes. We revealed that the imaging condition depends on the type of parametric process, namely, whether down- or up-conversion is used. Influence of information about sparsity in discrete cosine transform and Haar transform bases on reconstruction quality is studied. In addition, we compared ordinary and ghost images when the detectors are additionally illuminated by noise photons in a computer experiment, which showed increased noise immunity of GI, especially with processing via the proposed technique.