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

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