Lei et al. (2025) Balancing Accuracy and Efficiency: HWBENet for Water Body Extraction in Complex Rural Landscapes

⚠️ Warning: This summary was generated from the abstract only, as the full text was not available.

Identification

• Journal: Remote Sensing
• Year: 2025
• Date: 2025-11-14
• Authors: Pengyu Lei, Jiang Zhang, Jizheng Yi
• DOI: 10.3390/rs17223711

Research Groups

Not specified in the provided text.

Short Summary

This paper introduces HWBENet, a novel hybrid deep learning network designed for efficient and accurate extraction of water bodies from high-resolution remote sensing imagery, particularly in complex rural landscapes, by balancing computational cost with segmentation precision.

Objective

• To develop an efficient, accurate, and scalable deep learning model for extracting small, scattered, and irregularly shaped water bodies from high-resolution remote sensing imagery in complex rural landscapes, addressing the trade-off between precision and computational overhead.

Study Configuration

• Spatial Scale: Regional to large-scale areas, utilizing high-resolution remote sensing imagery.
• Temporal Scale: Implied for timely and large-scale monitoring applications.

Methodology and Data

• Models used: HWBENet, a novel hybrid network incorporating a lightweight MobileNetV3 encoder, a Contextual Information Mining Module (CIMM), and an Edge Refinement Module (ERM leveraging transformer mechanisms).
• Data sources: High-resolution remote sensing imagery; Challenging rural water body datasets.

Main Results

• HWBENet demonstrates a superior balance between accuracy and computational cost compared to existing state-of-the-art deep learning models.
• The network is validated as an efficient, accurate, and scalable solution for water body extraction.
• It offers significant practical value for large-scale hydrological mapping in complex rural environments.

Contributions

• Introduction of HWBENet, a novel hybrid network specifically designed to balance precision and efficiency for water body extraction.
• Development of the Contextual Information Mining Module (CIMM) to effectively fuse global context and fine-grained local details for fragmented water bodies.
• Integration of an Edge Refinement Module (ERM) utilizing transformer mechanisms to sharpen boundary details.
• Provision of an efficient, accurate, and scalable solution for large-scale hydrological mapping in challenging rural environments.

Funding

Not specified in the provided text.

Citation

@article{Lei2025Balancing,
  author = {Lei, Pengyu and Zhang, Jiang and Yi, Jizheng},
  title = {Balancing Accuracy and Efficiency: HWBENet for Water Body Extraction in Complex Rural Landscapes},
  journal = {Remote Sensing},
  year = {2025},
  doi = {10.3390/rs17223711},
  url = {https://doi.org/10.3390/rs17223711}
}