Kernel Estimation Using Total Variation Guided GAN for Image Super-Resolution
Various super-resolution (SR) kernels in the degradation model deteriorate the performance of the SR algorithms, showing unpleasant artifacts in the output images. Hence, SR kernel estimation has been studied to improve the SR performance in several ways for more than a decade. In particular, a conv...
Main Authors: | , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2023-04-01
|
Series: | Sensors |
Subjects: | |
Online Access: | https://www.mdpi.com/1424-8220/23/7/3734 |
_version_ | 1797606926917304320 |
---|---|
author | Jongeun Park Hansol Kim Moon Gi Kang |
author_facet | Jongeun Park Hansol Kim Moon Gi Kang |
author_sort | Jongeun Park |
collection | DOAJ |
description | Various super-resolution (SR) kernels in the degradation model deteriorate the performance of the SR algorithms, showing unpleasant artifacts in the output images. Hence, SR kernel estimation has been studied to improve the SR performance in several ways for more than a decade. In particular, a conventional research named KernelGAN has recently been proposed. To estimate the SR kernel from a single image, KernelGAN introduces generative adversarial networks(GANs) that utilize the recurrence of similar structures across scales. Subsequently, an enhanced version of KernelGAN, named E-KernelGAN, was proposed to consider image sharpness and edge thickness. Although it is stable compared to the earlier method, it still encounters challenges in estimating sizable and anisotropic kernels because the structural information of an input image is not sufficiently considered. In this paper, we propose a kernel estimation algorithm called Total Variation Guided KernelGAN (TVG-KernelGAN), which efficiently enables networks to focus on the structural information of an input image. The experimental results show that the proposed algorithm accurately and stably estimates kernels, particularly sizable and anisotropic kernels, both qualitatively and quantitatively. In addition, we compared the results of the non-blind SR methods, using SR kernel estimation techniques. The results indicate that the performance of the SR algorithms was improved using our proposed method. |
first_indexed | 2024-03-11T05:24:42Z |
format | Article |
id | doaj.art-57f88f5fbcf1441d8181ce31fff0e2e4 |
institution | Directory Open Access Journal |
issn | 1424-8220 |
language | English |
last_indexed | 2024-03-11T05:24:42Z |
publishDate | 2023-04-01 |
publisher | MDPI AG |
record_format | Article |
series | Sensors |
spelling | doaj.art-57f88f5fbcf1441d8181ce31fff0e2e42023-11-17T17:36:57ZengMDPI AGSensors1424-82202023-04-01237373410.3390/s23073734Kernel Estimation Using Total Variation Guided GAN for Image Super-ResolutionJongeun Park0Hansol Kim1Moon Gi Kang2School of Electrical and Electronic Engineering, Yonsei University, Seoul 03722, Republic of KoreaSchool of Electrical and Electronic Engineering, Yonsei University, Seoul 03722, Republic of KoreaSchool of Electrical and Electronic Engineering, Yonsei University, Seoul 03722, Republic of KoreaVarious super-resolution (SR) kernels in the degradation model deteriorate the performance of the SR algorithms, showing unpleasant artifacts in the output images. Hence, SR kernel estimation has been studied to improve the SR performance in several ways for more than a decade. In particular, a conventional research named KernelGAN has recently been proposed. To estimate the SR kernel from a single image, KernelGAN introduces generative adversarial networks(GANs) that utilize the recurrence of similar structures across scales. Subsequently, an enhanced version of KernelGAN, named E-KernelGAN, was proposed to consider image sharpness and edge thickness. Although it is stable compared to the earlier method, it still encounters challenges in estimating sizable and anisotropic kernels because the structural information of an input image is not sufficiently considered. In this paper, we propose a kernel estimation algorithm called Total Variation Guided KernelGAN (TVG-KernelGAN), which efficiently enables networks to focus on the structural information of an input image. The experimental results show that the proposed algorithm accurately and stably estimates kernels, particularly sizable and anisotropic kernels, both qualitatively and quantitatively. In addition, we compared the results of the non-blind SR methods, using SR kernel estimation techniques. The results indicate that the performance of the SR algorithms was improved using our proposed method.https://www.mdpi.com/1424-8220/23/7/3734kernel estimationgenerative adversarial networkssuper-resolutionself-similaritytotal variationKernelGAN |
spellingShingle | Jongeun Park Hansol Kim Moon Gi Kang Kernel Estimation Using Total Variation Guided GAN for Image Super-Resolution Sensors kernel estimation generative adversarial networks super-resolution self-similarity total variation KernelGAN |
title | Kernel Estimation Using Total Variation Guided GAN for Image Super-Resolution |
title_full | Kernel Estimation Using Total Variation Guided GAN for Image Super-Resolution |
title_fullStr | Kernel Estimation Using Total Variation Guided GAN for Image Super-Resolution |
title_full_unstemmed | Kernel Estimation Using Total Variation Guided GAN for Image Super-Resolution |
title_short | Kernel Estimation Using Total Variation Guided GAN for Image Super-Resolution |
title_sort | kernel estimation using total variation guided gan for image super resolution |
topic | kernel estimation generative adversarial networks super-resolution self-similarity total variation KernelGAN |
url | https://www.mdpi.com/1424-8220/23/7/3734 |
work_keys_str_mv | AT jongeunpark kernelestimationusingtotalvariationguidedganforimagesuperresolution AT hansolkim kernelestimationusingtotalvariationguidedganforimagesuperresolution AT moongikang kernelestimationusingtotalvariationguidedganforimagesuperresolution |