Water index and Swin Transformer Ensemble (WISTE) for water body extraction from multispectral remote sensing images

Ma, Donghui1; Jiang, Liguang2; Li, Jie3; Shi, Yun4

2023-12-31

10.1080/15481603.2023.2251704

GISCIENCE & REMOTE SENSING   Impact Factor & Quartile

1548-1603

1943-7226

Automatic surface water body mapping using remote sensing technology is greatly meaningful for studying inland water dynamics at regional to global scales. Convolutional neural networks (CNN) have become an efficient semantic segmentation technique for the interpretation of remote sensing images. However, the receptive field value of a CNN is restricted by the convolutional kernel size because the network only focuses on local features. The Swin Transformer has recently demonstrated its outstanding performance in computer vision tasks, and it could be useful for processing multispectral remote sensing images. In this article, a Water Index and Swin Transformer Ensemble (WISTE) method for automatic water body extraction is proposed. First, a dual-branch encoder architecture is designed for the Swin Transformer, aggregating the global semantic information and pixel neighbor relationships captured by fully convolutional networks (FCN) and multihead self-attention. Second, to prevent the Swin Transformer from ignoring multispectral information, we construct a prediction map ensemble module. The predictions of the Swin Transformer and the Normalized Difference Water Index (NDWI) are combined by a Bayesian averaging strategy. Finally, the experimental results obtained on two distinct datasets demonstrate that the WISTE has advantages over other segmentation methods and achieves the best results. The method proposed in this study can be used for improving regional to continental surface water mapping and related hydrological studies.

Deep learning water body extraction Swin Transformer water index dual-encoder

SCI

English

Corresponding

This work was partially supported by the Shenzhen Key Laboratory of Precision Measurement and Early Warning Technology for Urban Environmental Health Risks (ZDSYS20220606100604008), SUSTech research start-up grants (Y01296129; Y01296229), the CRSRI Open Re[Y01296129] ; Shenzhen Key Laboratory of Precision Measurement and Early Warning Technology for Urban Environmental Health Risks["Y01296229","SN: CKWV20221009/KY"] ; SUSTech research start-up grants[42174045] ; CRSRI Open Research Program[41874012] ; null[ZDSYS20220606100604008]

Physical Geography ; Remote Sensing

Geography, Physical ; Remote Sensing

WOS:001057442300001

TAYLOR & FRANCIS LTD

Web of Science

1.Geovis Spatial Technol Co Ltd, Xian, Peoples R China
2.Southern Univ Sci & Technol, Sch Environm Sci & Engn, Shenzhen Key Lab Precis Measurement & Early Warnin, Shenzhen, Peoples R China
3.Xian Surveying & Mapping Inst, Xian, Peoples R China
4.Xian Univ Sci & Technol, Coll Geomat, Xian, Peoples R China

Ma, Donghui,Jiang, Liguang,Li, Jie,et al. Water index and Swin Transformer Ensemble (WISTE) for water body extraction from multispectral remote sensing images[J]. GISCIENCE & REMOTE SENSING,2023,60(1).

Ma, Donghui,Jiang, Liguang,Li, Jie,&Shi, Yun.(2023).Water index and Swin Transformer Ensemble (WISTE) for water body extraction from multispectral remote sensing images.GISCIENCE & REMOTE SENSING,60(1).

Ma, Donghui,et al."Water index and Swin Transformer Ensemble (WISTE) for water body extraction from multispectral remote sensing images".GISCIENCE & REMOTE SENSING 60.1(2023).