A Method for Extracting Lake Water Using ViTenc-UNet: Taking Typical Lakes on the Qinghai-Tibet Plateau as Examples

Abstract

As the lakes located in the Qinghai-Tibet Plateau are important carriers of water resources in Asia, dynamic changes to these lakes intuitively reflect the climate and water resource variations of the Qinghai-Tibet Plateau. To address the insufficient performance of the Convolutional Neural Network (CNN) in learning the spatial relationship between long-distance continuous pixels, this study proposes a water recognition model for lakes on the Qinghai-Tibet Plateau based on U-Net and ViTenc-UNet. This method uses Vision Transformer (ViT) to replace the continuous Convolutional Neural Network layer in the encoder of the U-Net model, which can more accurately identify and extract the continuous spatial relationship of lake water bodies. A Convolutional Block Attention Module (CBAM) mechanism was added to the decoder of the model enabling the spatial information and spectral information characteristics of the water bodies to be more completely preserved. The experimental results show that the ViTenc-UNet model can complete the task of lake water recognition on the Qinghai-Tibet Plateau more efficiently, and the Overall Accuracy, Intersection over Union, Recall, Precision, and F1 score of the classification results for lake water bodies reached 99.04%, 98.68%, 99.08%, 98.59%, and 98.75%, which were, respectively, 4.16%, 6.20% 5.34%, 4.80%, and 5.34% higher than the original U-Net model. Compared to FCN, the DeepLabv3+, TransUNet, and Swin-Unet models also have different degrees of advantages. This model innovatively introduces ViT and CBAM into the water extraction task of lakes on the Qinghai-Tibet Plateau, showing excellent water classification performance of these lake bodies. This method has certain classification advantages and will provide an important scientific reference for the accurate real-time monitoring of important water resources on the Qinghai-Tibet Plateau.

Combining Satellite Images and the Hydraulic Engineering Archive to Map the Processes of Reservoir Construction in Xinjiang

“…U-Net is one of the most common CNN models for water body mapping in various scenarios [38][39][40], especially for small water body mapping on high-resolution images [27,41,42]. In this study, we used the U-Net model to map all the water bodies in Xinjiang from 2392 Sentinel-2 images from April to October 2022.…”

GEA-MSNet: A Novel Model for Segmenting Remote Sensing Images of Lakes Based on the Global Efficient Attention Module and Multi-Scale Feature Extraction

“…Furthermore, Zhang et al [27] proposed a U-Net network that integrates transformer modules, utilizing CONV+Mixformer Block for feature extraction and incorporating an attention module for optimal water segmentation performance on the GID datasets. Zhao et al [28] combined the swin-transformer algorithm for lake extraction on the Qinghai-Tibet Plateau lake datasets, replaced the convolution structure of the U-Net encoder with vision transformer (VIT) and added channel-wise and spatial-wise attention module (CBAM) attention mechanism in the decoder. The improved network achieved the best performance compared with other algorithms in the plateau lake extraction task.…”

An Enhanced Dual-Stream Network Using Multi-Source Remote Sensing Imagery for Water Body Segmentation

“…However, the application of deep learning in water bodies extraction faces several challenges, including diverse morphologies, complex scenarios (e.g., rivers, lakes, urban water bodies, and aquaculture), and variations in dimensions, with particular difficulty presented by small water bodies [8]. Convolutional Neural Network (CNN) have emerged as a prevalent strategy in water bodies extraction research, owing to their proficiency in semantic information of remote sensing images through convolutional operations processes [8,[10][11][12]. Despite their merits, CNNs are constrained by their receptive field dimensions, leading to a predominant focus on local features while ignoring the spatial correlation within a water body.…”

“…Despite their merits, CNNs are constrained by their receptive field dimensions, leading to a predominant focus on local features while ignoring the spatial correlation within a water body. This limitation often culminates in inaccurate edges of lake boundaries [10,11]. Furthermore, the progressive enlargement of the receptive field amidst continuous down-sampling can result in the neglect of smaller water bodies' features, leading to fragmented extraction outcomes [12].…”

“…Unlike traditional CNNs, a Transformer leverages multi-head attention to establish remote dependencies, thereby facilitating a more holistic feature extraction from images [20]. This technique is gaining traction in water bodies extraction [11,21]. Zhao et al [11] introduced Convolutional Block Attention Module (CBAM) and substituted the conventional convolutional layers in the original U-Net model encoder with Vision Transformer (Vi-T), enhancing the continuous spatial relationship interpretation, and thereby improving the segmentation of lake water bodies on the Qinghai-Tibet Plateau.…”