An optimal transport model for imaging in atmospheric turbulence

TL;DR

This paper introduces a physics-based optimal transport model for atmospheric turbulence imaging, improving the prediction and recovery of image distortions caused by turbulence.

Contribution

It develops a novel optimal transport framework linking photon flow and wave phase, enabling more accurate modeling of turbulence effects on images.

Findings

Model outperforms existing methods on simulated data

Validated with real atmospheric turbulence data

Suggests new algorithms for atmospheric imaging applications

Abstract

We describe a new model for image propagation through open air in the presence of changes in the index of refraction (e.g. due to turbulence) using the theory of optimal transport. We describe the relationship between photon density, or image intensity, and the phase of the traveling wave and, together with a least action principle, suggest a method for approximately recovering the solution of the photon flow. By linking atmospheric propagation solutions to optimal transport, we provide a physics-based (as opposed to phenomenological) model for predicting turbulence-induced changes to sequences of images. Simulated and real data are utilized to validate and compare the model to other existing methods typically used to model this type of data. Given its superior performance in describing experimental data, the new model suggests new algorithms for a variety of atmospheric imaging applications.