Evaluation of neural network algorithms for atmospheric turbulence mitigation

TL;DR

This paper reviews and experimentally evaluates various neural network architectures for mitigating atmospheric turbulence-induced blur in images and videos, focusing on reusability and architectural features.

Contribution

It provides a comparative analysis of five neural network architectures specifically tailored for atmospheric turbulence mitigation, highlighting their reusability and design aspects.

Findings

End-to-end trained network reduces the need for stabilization.

Certain architectures show better reusability for turbulence mitigation.

Comparison results guide future neural network design for atmospheric turbulence.

Abstract

A variety of neural networks architectures are being studied to tackle blur in images and videos caused by a non-steady camera and objects being captured. In this paper, we present an overview of these existing networks and perform experiments to remove the blur caused by atmospheric turbulence. Our experiments aim to examine the reusability of existing networks and identify desirable aspects of the architecture in a system that is geared specifically towards atmospheric turbulence mitigation. We compare five different architectures, including a network trained in an end-to-end fashion, thereby removing the need for a stabilization step.