Underwater Image Restoration With Light-Aware Progressive Network

Underwater images suffer from different types of degradation such as blur and color cast. It is difficult to obtain images with both color equalization and rich texture in various underwater scenes. To alleviate these issues, this paper proposes a reflected light-aware multi-scale progressive restoration network. A restoration framework is designed based on physical imaging model. It can accurately remove scattering blur of various underwater images and output a preliminary restoration result. A multi-scale progressive enhancement module is proposed to further enrich the image details and improve the generalization ability of the model by fully extracting detailed and spatial features. For the issues of color cast and low contrast, a new chromatic aberration correction mechanism is introduced, which makes the image color balanced. The experimental results show that our method can obtain underwater images with enriching texture and harmonious color distribution as compared with other state-of-the-art methods.