AUTHOR=Song Wei , Liu Yaling , Huang Dongmei , Zhang Bing , Shen Zhihao , Xu Huifang 

TITLE=From shallow sea to deep sea: research progress in underwater image restoration

JOURNAL=Frontiers in Marine Science

VOLUME=Volume 10 - 2023

YEAR=2023

URL=https://www.frontiersin.org/journals/marine-science/articles/10.3389/fmars.2023.1163831

DOI=10.3389/fmars.2023.1163831

ISSN=2296-7745

ABSTRACT=Underwater images play a vital role in various fields such as oceanographic engineering, marine exploitation, and marine environmental protection. However, the quality of underwater images is often severely degraded due to the complexities of the underwater environment and equipment limitations, which impedes relevant research advancement. As a result, underwater image restoration has become a significant research area in recent years. With the increasing interest in deep-sea exploration, deep-sea image restoration has emerged as a new research focus, presenting new challenges. This paper conducts a systematic review of underwater image restoration technology from the shallow sea to the deep sea, attempting to bridge the gap between the shallow-sea and deep-sea image restoration fields through experimental analysis. This paper first categorizes shallow-sea image restoration methods into three types: physical model-based methods, prior-based methods, and deep learning-based methods that integrate physical models and analyzes the core concepts and the characteristics of representative methods. Secondly, this paper summarizes the research status and primary challenges facing deep-sea image restoration, which include color cast and blur caused by underwater environmental characteristics and insufficient and uneven lighting caused by artificial light sources. Potential solutions are analyzed, including the application of general shallow-sea restoration methods to address color cast and blur, and using methods from related fields such as exposure image correction and low-light image enhancement to address lighting problems. Finally, comprehensive experiments are conducted to examine the feasibility of shallow-sea image restoration methods and some related image enhancement methods in the context of deep-sea image restoration. The experimental results provide insight into existing methods for solving the problem of deep-sea image restoration, and an in-depth discussion suggests several future development directions in deep-sea image restoration.