Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

IDA-UIE: An Iterative Framework for Deep Network-based Degradation Aware Underwater Image Enhancement

Pranjali Singh, Prithwijit Guha

TL;DR

This work proposes an iterative framework that individually identifies and resolves a dominant degradation condition and creates condition-specific datasets derived from high-quality images in two standard datasets, UIEB and EUVP.

Abstract

Underwater image quality is affected by fluorescence, low illumination, absorption, and scattering. Recent works in underwater image enhancement have proposed different deep network architectures to handle these problems. Most of these works have proposed a single network to handle all the challenges. We believe that deep networks trained for specific conditions deliver better performance than a single network learned from all degradation cases. Accordingly, the first contribution of this work lies in the proposal of an iterative framework where a single dominant degradation condition is identified and resolved. This proposal considers the following eight degradation conditions -- low illumination, low contrast, haziness, blurred image, presence of noise and color imbalance in three different channels. A deep network is designed to identify the dominant degradation condition. Accordingly, an appropriate deep network is selected for degradation condition-specific enhancement. The second contribution of this work is the construction of degradation condition specific datasets from good quality images of two standard datasets (UIEB and EUVP). This dataset is used to learn the condition specific enhancement networks. The proposed approach is found to outperform nine baseline methods on UIEB and EUVP datasets.

IDA-UIE: An Iterative Framework for Deep Network-based Degradation Aware Underwater Image Enhancement

TL;DR

This work proposes an iterative framework that individually identifies and resolves a dominant degradation condition and creates condition-specific datasets derived from high-quality images in two standard datasets, UIEB and EUVP.

Abstract

Underwater image quality is affected by fluorescence, low illumination, absorption, and scattering. Recent works in underwater image enhancement have proposed different deep network architectures to handle these problems. Most of these works have proposed a single network to handle all the challenges. We believe that deep networks trained for specific conditions deliver better performance than a single network learned from all degradation cases. Accordingly, the first contribution of this work lies in the proposal of an iterative framework where a single dominant degradation condition is identified and resolved. This proposal considers the following eight degradation conditions -- low illumination, low contrast, haziness, blurred image, presence of noise and color imbalance in three different channels. A deep network is designed to identify the dominant degradation condition. Accordingly, an appropriate deep network is selected for degradation condition-specific enhancement. The second contribution of this work is the construction of degradation condition specific datasets from good quality images of two standard datasets (UIEB and EUVP). This dataset is used to learn the condition specific enhancement networks. The proposed approach is found to outperform nine baseline methods on UIEB and EUVP datasets.
Paper Structure (54 sections, 30 equations, 48 figures, 23 tables)

This paper contains 54 sections, 30 equations, 48 figures, 23 tables.

Table of Contents

  1. Introduction
  2. Application
  3. Challenges
  4. Absorption and Scattering
  5. Suspended Particles
  6. Non-Uniform Illumination
  7. Fluorescence
  8. Major Contribution
  9. Related Work
  10. Traditional Methods
  11. Image Denoising
  12. Contrast Enhancement Techniques
  13. Color Correction Techniques
  14. Histogram Equalization Method
  15. CLAHE
  16. ...and 39 more sections

Figures (48)

  • Figure 1: Application areas of underwater image processing, highlighting its critical roles in marine life identification, oceanography, underwater archaeology, security and surveillance, robotics, photography and videography, mapping and navigation, and virtual reality tourism a3.
  • Figure 2: Challenges in underwater imaging include significant light attenuation due to absorption, scattering, and reflection by water molecules, suspended particles, and dissolved substances. The attenuation varies with wavelength, causing shorter wavelengths like blue and green to be absorbed and scattered more than longer wavelengths like red. Additionally, underwater images are affected by fluorescence, non-uniform illumination, and reduced visibility, making it essential to enhance image quality for better exploration and study of underwater environments a3.
  • Figure 3: Light attenuation in underwater environments, illustrating the exponential decay of light intensity due to absorption and scattering. The diagram shows how red light, with the longest wavelength, is absorbed first, while blue light, with the shortest wavelength, penetrates the farthest, resulting in a bluish tint in underwater images raveendran2021underwater
  • Figure 4: Absorption and scattering in underwater environments, showing how light interacts with floating particles. The diagram illustrates the effects of forward scattering and backward scattering on the visibility and clarity of underwater images raveendran2021underwater.
  • Figure 5: Non-uniform illumination and the presence of suspended particles in water, demonstrating how absorption and scattering lead to blurriness, reduced contrast, and loss of image quality. High turbidity and powerful artificial light sources exacerbate these effects, causing reflections and bright spots that obscure image details a5
  • ...and 43 more figures