Underwater image enhancement is a critical task for various marine applications, aiming to improve the visual quality of images captured in underwater environments. However, this task is fraught with challenges due to the distinct characteristics of underwater environments, such as varying light conditions, color distortion, and low visibility. This paper introduces DeepAquaEnhance, a cutting-edge framework designed to address the unique challenges of underwater image enhancement. Our approach incorporates a Hierarchical Multi-head Attention (HMA) mechanism and Adaptive Feature Integration (AFI) to effectively enhance image clarity, color balance, and contrast while preserving the intrinsic characteristics of underwater scenes. Additionally, the model incorporates adversarial training and a Transformer architecture, which collectively ensure robust performance across a wide range of underwater conditions. We rigorously evaluate DeepAquaEnhance on several benchmark datasets, including LSUI, UIEB, UFO, and EUVP. The results demonstrate that our model consistently outperforms existing state-of-the-art methods, such as CLAHE, Funie-GAN, Fusion, IBLA, UDCP, U-Transformer, P2C, and PUGAN. Notably, DeepAquaEnhance excels in preserving the natural appearance of underwater scenes while significantly enhancing their visual quality. These advancements underscore the potential of DeepAquaEnhance to set a new standard in underwater image enhancement, offering a reliable and effective solution for improving underwater imagery. Our findings indicate a promising direction for future research and development, paving the way for more sophisticated and versatile underwater imaging technologies. The code and results are available on the following GitHub link: https://github.com/Rehman1995/DeepAquaEnhance . • Propose DeepAquaEnhance, a deep learning framework designed to improve the quality of underwater images for marine remote sensing and geoinformatics applications. • Integrate Hierarchical Multi-head Attention (HMA) and Adaptive Feature Integration (AFI) to enhance spatial detail, color fidelity, and contrast in diverse underwater scenes. • Incorporate Transformer-based modules and adversarial training to achieve robust enhancement under varying optical and environmental underwater conditions. • Demonstrate superior performance on benchmark datasets (LSUI, UIEB, UFO, EUVP) compared to state-of-the-art enhancement methods. • Improve the reliability of underwater imagery for marine habitat mapping, underwater archeological surveys, and environmental monitoring, supporting data-driven decision-making in geospatial sciences.
Rehman et al. (Wed,) studied this question.