Objectives  Underwater image plays an extremely important role in underwater resource exploration and non-contact inspection of underwater pipelines. However, due to absorption and scattering by water and particles in the water, the phenomenon of low contrast and blurred edges are existed in underwater image. An image enhancement algorithm is proposed to reduce the influence of water medium on underwater image.

  Methods  Firstly, the mathematical model of underwater imaging was studied, and the non-degraded image mathematical model was deduced according to Lambert-Beer law. Secondly, the red dark channel transmittance image in the non-degraded image model was obtained based on the Red Dark Channel Prior (RDCP) theory, and the non-degraded image with high contrast was obtained by filtering and thinning the image with the guided filter algorithm. In addition, since the image contrast was still not ideal after deblurring, the Contrast Limited Adaptive Histogram Equalization (CLAHE) algorithm was employed to improve the image contrast.

  Results  Based on the commonly used image evaluation criteria, the effectiveness of the proposed algorithm in this paper is evaluated by experiments.

  Conclusions  It is found that the algorithm proposed in this paper can significantly improve the information entropy, contrast and average gradient of underwater images, and more information is obtained from the underwater image.