2nd Place Solutions for UG2+ Challenge 2022 -- D$^{3}$Net for Mitigating Atmospheric Turbulence from Images

TL;DR

This paper presents D$^{3}$Net, a novel method for atmospheric turbulence mitigation in images, achieving state-of-the-art results in the UG2+ Challenge 2022, especially for text recognition and image enhancement.

Contribution

Introduction of D$^{3}$Net, a new approach for atmospheric turbulence mitigation that outperforms existing methods in challenging scenarios.

Findings

Achieved state-of-the-art performance on textual and balloon images.

Ranked 2nd in the UG2+ Challenge 2022.

Visual comparisons show superior results over existing denoising and deblurring methods.

Abstract

This technical report briefly introduces to the D$^{3}$Net proposed by our team "TUK-IKLAB" for Atmospheric Turbulence Mitigation in $UG2^{+}$ Challenge at CVPR 2022. In the light of test and validation results on textual images to improve text recognition performance and hot-air balloon images for image enhancement, we can say that the proposed method achieves state-of-the-art performance. Furthermore, we also provide a visual comparison with publicly available denoising, deblurring, and frame averaging methods with respect to the proposed work. The proposed method ranked 2nd on the final leader-board of the aforementioned challenge in the testing phase, respectively.