我国海洋探索正向纵深发展。在深海探测中光学影像发挥着至关重要的“眼睛”作用，但由于受深海探测环境的影响，高质量的图像获取仍面临严竣挑战。本项目基于对深海探测的成像环境感知开展深海图像增强方法研究，针对深海图像的真实感弱、人工光源影响性强、易受深海环境干扰的特点，重点研究1）基于环境感知的深海光学成像模型；2）深海图像的复原和增强方法；3）深海复杂环境自适应的图像实时优化策略。通过解决研究中的关键技术问题，包括：深海光学成像模型的理论构建、基于成像模型的深海图像光吸收补偿、不均匀光照的复原方法、基于学习的环境自适应增强策略设计，实现基于图像的后向深海成像环境感知，能够对千米到成万米深海探测中的图像进行有效增强，并在深海模拟实验场和“探索一号”深海探测任务中对所提出的理论和方法进行试验和验证。研究成果形成深海图像增强的理论体系与系统，为提高深海探测效率、节约成本、促进海洋科研发展提供应用基础。

With the development of China's deep-sea exploration, optical imaging plays an increasingly important role. However, due to the influence of deep-sea environment, the acquisition of high-quality deep-sea images still faces severe challenges. This project carries out the research on deep-sea image enhancement based on the sensing of imaging environment in deep sea exploration. In view of the characteristics of deep-sea images of the weak realities, the strong influence by artificial light, the effect by the deep-sea environment, this research focuses on 1) environmental perception based deep-sea optical imaging model; 2) deep-sea image restoration and enhancement method; 3) adaptive image optimization strategy for deep-sea complex environment. By tackling key technologies in the study of the problems, including: the theory of optical imaging model in deep sea, the method of deep light absorption compensation, the method of non-uniform illumination image restoration, the design of the environment adaptive enhancement strategy, we aim to aware the imaging environment from the deep sea image only, and to enhance effectively the images from tens of and thousands of kilometers deep-sea exploration. The proposed theory and methods will be tested and verified in deep sea simulation experiment field and in the “Exploration 1” tasks. The research outcome will form the theoretical and system of deep-sea image enhancement, and provide the application basis for improving the efficiency of deep-sea exploration, saving costs and promoting the development of Marine scientific research.