Related Experiment Video
Updated: Jul 8, 2026
03:31
Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
Published on: December 15, 2023
533
Underwater image enhancement using Divide-and-Conquer network.
Shijian Zheng1,2, Rujing Wang2,3, Guo Chen1
1Department of Information Engineering, Southwest University of Science and Technology, Mianyang, Sichuan, China.
Plos One
|March 5, 2024
Summary
This study introduces a novel Divide-and-Conquer network (DC-net) for enhancing underwater images. The DC-net effectively addresses color distortion and blur, improving visual quality and information extraction from underwater scenes.
Area of Science:
- Computer Vision
- Image Processing
- Artificial Intelligence
Background:
- Underwater images suffer from color distortion and blur due to light attenuation and water conditions.
- Existing underwater image enhancement methods often struggle with comprehensive quality improvement.
Purpose of the Study:
- To develop an effective method for enhancing underwater images.
- To improve both the visual quality and the information extractability of underwater imagery.
Main Methods:
- A novel Divide-and-Conquer network (DC-net) was designed.
- The DC-net integrates a texture network with multi-axis attention, a color network using adaptive 3D look-up tables, and a refinement network.
Main Results:
- The proposed DC-net achieved superior visual quality in underwater images.
- Qualitative and quantitative performance surpassed state-of-the-art underwater image enhancement methods.
Conclusions:
- The DC-net offers a robust solution for underwater image enhancement.
- The method demonstrates significant improvements in handling complex underwater image degradations.
