Abstract:Accurately mapping the surface water bodies through remote sensing technology is of great significance for water resources management, flood monitoring, and drought monitoring. At present, many scholars at home and abroad carry out research on deep learning image recognition algorithms based on convolutional neural networks, and a variety of variant-based convolutional neural networks are proposed to be applied to extract water bodies from remote sensing images. However, due to the low depth of convolutional layers employed and underutilization of water spectral feature information, most of the water body extraction methods based on convolutional neural networks (CNNs) for remote sensing images are limited in accuracy. In this study, we propose a novel surface water automatic extraction method based on the convolutional neural network (CRAUnet++) for Sentinel-2 images. The proposed method includes three parts: (1) substituting the feature extractor of the original Unet++ with ResNet34 to enhance the network's complexity by increasing its depth; (2) Embedding the Spatial and Channel 'Squeeze and Excitation' (SCSE) module into the up-sampling stage of the network to suppress background features and amplify water body features; (3) adding the vegetation red edge-based water index (RWI) into the input data to maximize the utilization of water body spectral information of Sentinel-2 images without increasing the data processing time. To verify the performance and accuracy of the proposed algorithm, the ablation experiment under four different strategies and comparison experiment with different algorithms of RWI, FCN, SegNet, Unet, and DeepLab v3+ were conducted on Sentinel-2 images of the Poyang Lake. The experimental result shows that the precision, recall, F1, and IoU of CRAUnet++ are 95.99%, 96.41%, 96.19%, and 92.67%, respectively. CRAUnet++ has a good performance in extracting various types of water bodies and suppressing noises because it introduces SCSE attention mechanisms and combines surface water spectral features from RWI, exceeding that of the other five algorithms. The result demonstrates that CRAUnet++ has high validity and reliability in extracting surface water bodies based on Sentinel-2 images.

MST-UNet: a modified Swin Transformer for water bodies' mapping using Sentinel-2 images

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

Attention Swin Transformer UNet for Landslide Segmentation in Remotely Sensed Images

STIRUnet: SwinTransformer and inverted residual convolution embedding in unet for Sea-Land segmentation

SPT-UNet: A Superpixel-Level Feature Fusion Network for Water Extraction from SAR Imagery

SwinDefNet: A Novel Surface Water Mapping Model in Mountain and Cloudy Regions Based on Sentinel-2 Imagery

TSAE-UNet: A Novel Network for Multi-Scene and Multi-Temporal Water Body Detection Based on Spatiotemporal Feature Extraction

A Siamese Swin-Unet for image change detection

MRSE-Net: Multiscale Residuals and SE-Attention Network for Water Body Segmentation From Satellite Images

A Semantic Segmentation Method for Remote Sensing Images Based on the Swin Transformer Fusion Gabor Filter

SwinSUNet: Pure Transformer Network for Remote Sensing Image Change Detection

UNetFormer: A UNet-like transformer for efficient semantic segmentation of remote sensing urban scene imagery

CRAUnet++: A New Convolutional Neural Network for Land Surface Water Extraction from Sentinel-2 Imagery by Combining RWI with Improved Unet++

P-Swin: Parallel Swin transformer multi-scale semantic segmentation network for land cover classification

MCAT-UNet: Convolutional and Cross-Shaped Window Attention Enhanced UNet for Efficient High-Resolution Remote Sensing Image Segmentation

TMNet: A Two-Branch Multi-Scale Semantic Segmentation Network for Remote Sensing Images

Local-enhanced multi-scale aggregation swin transformer for semantic segmentation of high-resolution remote sensing images

DS-TransUNet:Dual Swin Transformer U-Net for Medical Image Segmentation

Improvement of deep learning Method for water body segmentation of remote sensing images based on attention modules

Water body extraction from high spatial resolution remote sensing images based on enhanced U-Net and multi-scale information fusion