Artificial light sources are necessary to assist the illumination especially for underwater environment with dim natural light. However, inappropriate artificial lighting will lead to uneven brightness of the picture. The quality of underwater images taken by this hybrid light source is severely degraded, which not only affects visual perception, but also poses a challenge to the development of subsequent high-level computer vision tasks. However, most of the existing methods only consider the influence of natural light source, which is not good for underwater image restoration under hybrid light source environment. To handle the problems of uneven brightness, color bias and blurry details, we propose an Illumination-aware Encoder-Decoder Network (IEDN) with attention mechanism for underwater image restoration. On the one hand, attention mechanism and enhanced residual block are incorporated into a multi-scale structure to effectively extract detailed structure features. On the other hand, illumination aware map is introduced as a priory constraint to balance the contrast of the restoration results. Meanwhile, appropriate loss functions are designed to guide the network to fully learn the nonlinear mapping relationship between the underwater image and the ground truth. In this way, the tone of the restored image is more natural and the texture details are more plentiful. The results of the comparison experiments prove that the proposed method is superior to other algorithms both quantitively and qualitatively. The ablation experiments further demonstrate the effectiveness of the network modules and the illumination aware.
