Underwater Image Restoration via Non-Convex Non-Smooth Variation and Thermal Exchange Optimization

Journal of Marine Science and Engineering

Ming Liu¹  Yuejin Zhao¹  Qingliang Jiao¹  Pengyu Li¹  Mei Hui¹  Lingqin Kong¹  Liquan Dong¹ 

[1] Beijing Key Laboratory for Precision Optoelectronic Measurement Instrument and Technology, School of Optics and Photonics, Beijing Institute of Technology, Beijing 100081, China;

关键词: underwater image restoration;    underwater DCP (dark channel prior);    non-convex non-smooth variation;    adapt parameter selection;    thermal exchange optimization;   

DOI  :  10.3390/jmse9060570

来源: DOAJ

【 摘 要 】

The quality of underwater images is an important problem for resource detection. However, the light scattering and plankton in water can impact the quality of underwater images. In this paper, a novel underwater image restoration based on non-convex, non-smooth variation and thermal exchange optimization is proposed. Firstly, the underwater dark channel prior is used to estimate the rough transmission map. Secondly, the rough transmission map is refined by the proposed adaptive non-convex non-smooth variation. Then, Thermal Exchange Optimization is applied to compensate for the red channel of underwater images. Finally, the restored image can be estimated via the image formation model. The results show that the proposed algorithm can output high-quality images, according to qualitative and quantitative analysis.

【 授权许可】

Unknown