Applications of Deep Learning-Based Super-Resolution for Sea Surface Temperature Reconstruction

doi:10.1109/JSTARS.2020.3042242

	Applications of Deep Learning-Based Super-Resolution for Sea Surface Temperature Reconstruction
	Ping, Bo 1; Su, Fenzhen 2; Han, Xingxing 1; Meng, Yunshan 3
刊名	IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING
	2021
卷号	14 页码:887-896
关键词	Advanced microwave scanning radiometer 2 (AMSR2) deep learning moderate-resolution imaging spectroradiometer (MODIS) sea surface temperature super-resolution
ISSN号	1939-1404
DOI	10.1109/JSTARS.2020.3042242
通讯作者	Ping, Bo(pingbo@tju.edu.cn)
英文摘要	Deep learning-based super-resolution (SR) methods have been widely used in natural images; however, their applications in satellite-derived sea surface temperature (SST) have not yet been fully discussed. Hence, it is necessary to analyze the validity of deep learning-based SR methods in SST reconstruction. In this study, an SR model, including multiscale feature extraction and multireceptive field mapping, was first proposed. Then, the proposed model and four other existing SR models were applied to SST reconstruction and analyzed. First, compared with the bicubic interpolation method, the SR models can improve the reconstruction accuracy. Compared with four other SR models, the proposed model can achieve the lowest mean squared error (MAE) in the East China Sea (ECS), in the northwest Pacific (NWP) and in the west Atlantic (WA), the second-lowest MAE in the southeast Pacific (SEP); the lowest root mean squared error (RMSE) in ECS andWA, the second-lowest RMSE in NWP and SEP. Additionally, ODRE model can acquire the highest or the second-highest peak single-to-noise ratio and structural similarity index in ECS, NWP, and SEP. Moreover, the number of missing pixels and SST variety are two essential factors in the SRperformance. The proposed multiscale feature extraction process can enhance the SR performance, especially for small regions and stable SST regions. Finally, while a deeper network can be helpful in achieving SR performance, the approach of simply adding more dilation convolutions may not enhance the reconstruction accuracy.
资助项目	Natural Science Foundation of Tianjin[18JCQNJC08900] ; State Key Laboratory of Resources and Environmental Information System
WOS研究方向	Engineering ; Physical Geography ; Remote Sensing ; Imaging Science & Photographic Technology
语种	英语
出版者	IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
WOS记录号	WOS:000640582500001
资助机构	Natural Science Foundation of Tianjin ; State Key Laboratory of Resources and Environmental Information System
内容类型	期刊论文
源URL	[http://ir.igsnrr.ac.cn/handle/311030/161680]
专题	中国科学院地理科学与资源研究所
通讯作者	Ping, Bo
作者单位	1.Tianjin Univ, Inst Surface Earth Syst Sci, Sch Earth Syst Sci, Tianjin 300072, Peoples R China 2.Univ Chinese Acad Sci, Inst Geog Sci & Nat Resources Res, LREIS, Beijing 100101, Peoples R China 3.Natl Marine Data & Informat Serv, Tianjin 300171, Peoples R China
推荐引用方式 GB/T 7714	Ping, Bo,Su, Fenzhen,Han, Xingxing,et al. Applications of Deep Learning-Based Super-Resolution for Sea Surface Temperature Reconstruction[J]. IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING,2021,14:887-896.
APA	Ping, Bo,Su, Fenzhen,Han, Xingxing,&Meng, Yunshan.(2021).Applications of Deep Learning-Based Super-Resolution for Sea Surface Temperature Reconstruction.IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING,14,887-896.
MLA	Ping, Bo,et al."Applications of Deep Learning-Based Super-Resolution for Sea Surface Temperature Reconstruction".IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING 14(2021):887-896.