Related Experiment Video
Updated: Sep 2, 2025
03:31
Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
Published on: December 15, 2023
623
New deep learning method for efficient extraction of small water from remote sensing images
Yuanjiang Luo1, Ao Feng1, Hongxiang Li1
1College of Information Engineering, Sichuan Agricultural University, Ya'an, Sichuan, China.
Plos One
|August 5, 2022
Summary
This study introduces a new method for extracting water bodies from remote sensing images, improving accuracy for small water features. The enhanced approach achieves 94.72% prediction accuracy, outperforming existing techniques.
Area of Science:
- Remote Sensing
- Geospatial Analysis
- Image Processing
Background:
- Accurate water body extraction from remote sensing data is crucial for water resource management.
- Traditional methods and existing neural networks face challenges in precisely delineating fine water bodies and maintaining overall accuracy.
- Existing approaches often require multiple methods or struggle with small, discrete water features.
Purpose of the Study:
- To develop an improved method for accurate water body extraction from remote sensing images, particularly focusing on fine and discrete water bodies.
- To enhance the feature representation of tiny water bodies and reduce the computational cost of training.
- To evaluate the proposed method against conventional and deep learning-based approaches.
Main Methods:
- Utilized false color processing and a generative adversarial network (GAN) to reconstruct and enhance remote sensing images, emphasizing small water features.
- Implemented a multi-scale input strategy to decrease training costs.
- Adapted the DeepLabv3+ architecture by incorporating strip pooling for improved extraction of discrete, long-distance water bodies.
Main Results:
- The proposed method demonstrated improved accuracy in water body extraction, proving effective for fine water bodies.
- Achieved a prediction accuracy of 94.72%, surpassing seven traditional and deep learning-based semantic segmentation methods.
- The strip pooling mechanism effectively captured water bodies with discrete distributions at long distances.
Conclusions:
- The novel water body extraction method, enhanced by GANs and strip pooling, offers superior performance compared to existing techniques.
- The approach successfully addresses the limitations of previous methods in extracting fine and discrete water bodies.
- This research contributes a more accurate and efficient solution for water body extraction in remote sensing applications.
Related Concept Videos
00:56Extraction: Advanced Methods
519
Metal ions can be separated from one another by complexation with organic ligands–the chelating agent– to form uncharged chelates. Here, the chelating agent must contain hydrophobic groups and behave as a weak acid, losing a proton to bind with the metal. Since most organic ligands used in this process are insoluble or undergo oxidation in the aqueous phase, the chelating agent is initially added to the organic phase and extracted into the aqueous phase. The metal-ligand complex is...
519
01:18Uniform Depth Channel Flow: Problem Solving
122
To calculate the flow rate for a trapezoidal channel, first, identify the bottom width, side slope, and flow depth of the channel. The cross-sectional area (A) corresponding to the depth of flow (y), channel bottom width (B), and side slope (θ) is determined by:Next, calculate the wetted perimeter, which includes the bottom width and the sloped side lengths in contact with the water. Using the values of the cross-sectional area and the wetted perimeter, determine the hydraulic radius by...
122