Super-resolution of sea surface temperature with convolutional neural network- and generative adversarial network-based methods
In this paper, we perform the super-resolution of sea surface temperature data with the enhanced super-resolution generative adversarial network (ESRGAN), which is a deep neural network-based single-image super-resolution (SISR) method that uses a generative adversarial network (GAN). We generate hi...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IWA Publishing
2022-04-01
|
| Series: | Journal of Water and Climate Change |
| Subjects: | |
| Online Access: | http://jwcc.iwaponline.com/content/13/4/1673 |
| _version_ | 1818140948277755904 |
|---|---|
| author | Tomoki Izumi Motoki Amagasaki Kei Ishida Masato Kiyama |
| author_facet | Tomoki Izumi Motoki Amagasaki Kei Ishida Masato Kiyama |
| author_sort | Tomoki Izumi |
| collection | DOAJ |
| description | In this paper, we perform the super-resolution of sea surface temperature data with the enhanced super-resolution generative adversarial network (ESRGAN), which is a deep neural network-based single-image super-resolution (SISR) method that uses a generative adversarial network (GAN). We generate high-quality super-resolution data with ESRGAN and with the super-resolution convolutional neural network (SRCNN) and residual-in-residual dense block network (RRDBNet) methods, which are based on convolutional neural networks (CNNs). The images generated with these methods are compared with high-resolution optimum interpolation sea surface temperature (OISST) data using root mean square error (RMSE), learned perceptual image patch similarity (LPIPS), and perceptual index (PI) evaluation methods. RRDBNet has a better RMSE than SRCNN and ESRGAN. However, CNN-based SISR methods do not provide a faithful representation of the ocean currents of OISST. ESRGAN has a better LPIPS and PI than CNN-based methods and can represent the complex distribution of ocean currents. HIGHLIGHTS
RRDBNet has a better RMSE than SRCNN and ESRGAN on super-resolution of sea surface temperature data.;
ESRGAN has a better LPIPS and PI than CNN-based methods and can represent the complex distribution of ocean currents.;
CNNs cannot interpolate the missing information, but GANs have better results for these parts.; |
| first_indexed | 2024-12-11T10:52:05Z |
| format | Article |
| id | doaj.art-eb47e9bff0e94a3ead7855f685bc6bf0 |
| institution | Directory Open Access Journal |
| issn | 2040-2244 2408-9354 |
| language | English |
| last_indexed | 2024-12-11T10:52:05Z |
| publishDate | 2022-04-01 |
| publisher | IWA Publishing |
| record_format | Article |
| series | Journal of Water and Climate Change |
| spelling | doaj.art-eb47e9bff0e94a3ead7855f685bc6bf02022-12-22T01:10:15ZengIWA PublishingJournal of Water and Climate Change2040-22442408-93542022-04-011341673168310.2166/wcc.2022.291291Super-resolution of sea surface temperature with convolutional neural network- and generative adversarial network-based methodsTomoki Izumi0Motoki Amagasaki1Kei Ishida2Masato Kiyama3 Graduate School of Science and Technology, Kumamoto University, 2-39-1 Kurokami, Chuo-ku, Kumamoto 860-8555, Japan Faculty of Advanced Science and Technology, Kumamoto University, 2-39-1 Kurokami, Chuo-ku, Kumamoto 860-8555, Japan Faculty of Advanced Science and Technology, Kumamoto University, 2-39-1 Kurokami, Chuo-ku, Kumamoto 860-8555, Japan Faculty of Advanced Science and Technology, Kumamoto University, 2-39-1 Kurokami, Chuo-ku, Kumamoto 860-8555, Japan In this paper, we perform the super-resolution of sea surface temperature data with the enhanced super-resolution generative adversarial network (ESRGAN), which is a deep neural network-based single-image super-resolution (SISR) method that uses a generative adversarial network (GAN). We generate high-quality super-resolution data with ESRGAN and with the super-resolution convolutional neural network (SRCNN) and residual-in-residual dense block network (RRDBNet) methods, which are based on convolutional neural networks (CNNs). The images generated with these methods are compared with high-resolution optimum interpolation sea surface temperature (OISST) data using root mean square error (RMSE), learned perceptual image patch similarity (LPIPS), and perceptual index (PI) evaluation methods. RRDBNet has a better RMSE than SRCNN and ESRGAN. However, CNN-based SISR methods do not provide a faithful representation of the ocean currents of OISST. ESRGAN has a better LPIPS and PI than CNN-based methods and can represent the complex distribution of ocean currents. HIGHLIGHTS RRDBNet has a better RMSE than SRCNN and ESRGAN on super-resolution of sea surface temperature data.; ESRGAN has a better LPIPS and PI than CNN-based methods and can represent the complex distribution of ocean currents.; CNNs cannot interpolate the missing information, but GANs have better results for these parts.;http://jwcc.iwaponline.com/content/13/4/1673convolutional neural networkesrgangenerative adversarial networkrrdbnetsingle-image super-resolution |
| spellingShingle | Tomoki Izumi Motoki Amagasaki Kei Ishida Masato Kiyama Super-resolution of sea surface temperature with convolutional neural network- and generative adversarial network-based methods Journal of Water and Climate Change convolutional neural network esrgan generative adversarial network rrdbnet single-image super-resolution |
| title | Super-resolution of sea surface temperature with convolutional neural network- and generative adversarial network-based methods |
| title_full | Super-resolution of sea surface temperature with convolutional neural network- and generative adversarial network-based methods |
| title_fullStr | Super-resolution of sea surface temperature with convolutional neural network- and generative adversarial network-based methods |
| title_full_unstemmed | Super-resolution of sea surface temperature with convolutional neural network- and generative adversarial network-based methods |
| title_short | Super-resolution of sea surface temperature with convolutional neural network- and generative adversarial network-based methods |
| title_sort | super resolution of sea surface temperature with convolutional neural network and generative adversarial network based methods |
| topic | convolutional neural network esrgan generative adversarial network rrdbnet single-image super-resolution |
| url | http://jwcc.iwaponline.com/content/13/4/1673 |
| work_keys_str_mv | AT tomokiizumi superresolutionofseasurfacetemperaturewithconvolutionalneuralnetworkandgenerativeadversarialnetworkbasedmethods AT motokiamagasaki superresolutionofseasurfacetemperaturewithconvolutionalneuralnetworkandgenerativeadversarialnetworkbasedmethods AT keiishida superresolutionofseasurfacetemperaturewithconvolutionalneuralnetworkandgenerativeadversarialnetworkbasedmethods AT masatokiyama superresolutionofseasurfacetemperaturewithconvolutionalneuralnetworkandgenerativeadversarialnetworkbasedmethods |