| Underwater images are important carriers of ocean information,which are widely used in fields such as marine military,marine environmental protection,and marine engineering.However,due to the attenuation and scattering properties of light,underwater images are generally subject to color deviations and low contrast,which is not conducive to subsequent research and exploration of ocean resources.To alleviate these two problems,research is conducted on underwater image restoration and enhancement algorithms,and completes the following works:(1)To reflect the underwater scenes better,the model parameter estimation methods are improved based on the underwater imaging model,including low-pass filter-based ambient light estimation,red channel inversion-based transmission derivation,saturation detectionbased transmission integration,and the water type-based transmission correction.Experimental results show that considering the influence of light’s attenuation characteristics,artificial light,ambient light,and water type can restore underwater images with higher visual quality.(2)To make the deep learning model more suitable for underwater image enhancement tasks,a multi-color and multi-stage collaborative network according to the degradation characteristic of color deviations of underwater images is proposed.For improving the model’s understanding of underwater scenes,a transmission guidance block is introduced to embed physical knowledge into the deep learning model,and proposes a physical-inconsistency loss to guide the deep learning model to learn physical prior knowledge.Extensive experiments show that the organic combination of physical and deep learning models can complete underwater image enhancement tasks better in different scenes.(3)The proposed underwater image restoration and enhancement algorithms have achieved the best comprehensive performance on public datasets,proving the effectiveness of the proposed algorithm.(4)After completing the algorithm research of underwater image enhancement,a simple implementation of the underwater image enhancement system is carried out,enabling users to obtain processing results with good visual quality on the system. |
