# Deep Underwater Image Enhancement

arXiv: https://arxiv.org/abs/1807.03528 · HTML: https://mlsanity.com/paper/1807.03528/

## Key facts

- Cited by 24 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 10 Jul 2018.

## Built on by (most cited first, 24 of 24)

- [Wavelet-based Fourier Information Interaction with Frequency Diffusion Adjustment for Underwater Image Restoration](https://mlsanity.com/paper/2311.16845/) (2023)
- [UIEC^2-Net: CNN-based Underwater Image Enhancement Using Two Color Space](https://mlsanity.com/paper/2103.07138/) (2021)
- [Toward Sufficient Spatial-Frequency Interaction for Gradient-aware Underwater Image Enhancement](https://mlsanity.com/paper/2309.04089/) (2023)
- [Learning A Physical-aware Diffusion Model Based on Transformer for Underwater Image Enhancement](https://mlsanity.com/paper/2403.01497/) (2024)
- [Underwater Light Field Retention : Neural Rendering for Underwater Imaging](https://mlsanity.com/paper/2203.11006/) (2022)
- [Low-Light Image and Video Enhancement: A Comprehensive Survey and Beyond](https://mlsanity.com/paper/2212.10772/) (2022)
- [A Wavelet-based Dual-stream Network for Underwater Image Enhancement](https://mlsanity.com/paper/2202.08758/) (2022)
- [Underwater Image Enhancement via Learning Water Type Desensitized Representations](https://mlsanity.com/paper/2102.00676/) (2021)
- [Investigation of the Challenges of Underwater-Visual-Monocular-SLAM](https://mlsanity.com/paper/2306.08738/) (2023)
- [PHISWID: Physics-Inspired Underwater Image Dataset Synthesized from RGB-D Images](https://mlsanity.com/paper/2404.03998/) (2024)
- [Marine Snow Removal Benchmarking Dataset](https://mlsanity.com/paper/2103.14249/) (2021)
- [A comprehensive benchmark analysis for sand dust image reconstruction](https://mlsanity.com/paper/2202.03031/) (2022)
- [DPF-Net: Physical Imaging Model Embedded Data-Driven Underwater Image Enhancement](https://mlsanity.com/paper/2503.12470/) (2025)
- [Global Modeling Matters: A Fast, Lightweight and Effective Baseline for Efficient Image Restoration](https://mlsanity.com/paper/2507.13663/) (2025)
- [A Survey of Fish Tracking Techniques Based on Computer Vision](https://mlsanity.com/paper/2110.02551/) (2021)
- [A deep learning approach for marine snow synthesis and removal](https://mlsanity.com/paper/2311.15584/) (2023)
- [A 7K Parameter Model for Underwater Image Enhancement based on Transmission Map Prior](https://mlsanity.com/paper/2405.16197/) (2024)
- [Advanced Underwater Image Quality Enhancement via Hybrid Super-Resolution Convolutional Neural Networks and Multi-Scale Retinex-Based Defogging Techniques](https://mlsanity.com/paper/2410.14285/) (2024)
- [A high-precision underwater object detection based on joint self-supervised deblurring and improved spatial transformer network](https://mlsanity.com/paper/2203.04822/) (2022)
- [Virtual Underwater Datasets for Autonomous Inspections](https://mlsanity.com/paper/2209.06013/) (2022)
- [An Adaptive Underwater Image Enhancement Framework via Multi-Domain Fusion and Color Compensation](https://mlsanity.com/paper/2503.03640/) (2025)
- [FUSION: Frequency-guided Underwater Spatial Image recOnstructioN](https://mlsanity.com/paper/2504.01243/) (2025)
- [Beyond Aesthetics: Quantifying Information Loss in Turbid Scenes](https://mlsanity.com/paper/2606.26295/) (2026)
- [Enhancing Underwater Images Using Deep Learning with Subjective Image Quality Integration](https://mlsanity.com/paper/2507.05393/) (2025)

## Abstract (arXiv)

In an underwater scene, wavelength-dependent light absorption and scattering degrade the visibility of images, causing low contrast and distorted color casts. To address this problem, we propose a convolutional neural network based image enhancement model, i.e., UWCNN, which is trained efficiently using a synthetic underwater image database. Unlike the existing works that require the parameters of underwater imaging model estimation or impose inflexible frameworks applicable only for specific scenes, our model directly reconstructs the clear latent underwater image by leveraging on an automatic end-to-end and data-driven training mechanism. Compliant with underwater imaging models and optical properties of underwater scenes, we first synthesize ten different marine image databases. Then, we separately train multiple UWCNN models for each underwater image formation type. Experimental results on real-world and synthetic underwater images demonstrate that the presented method generalizes well on different underwater scenes and outperforms the existing methods both qualitatively and quantitatively. Besides, we conduct an ablation study to demonstrate the effect of each component in our network.