Abstract There are some shortcomings of low resolution, color slants and low lightness in underwater image. Thus, this seriously affects the ability of underwater robot. This paper studies on the visual enhancement problem of robot, and proposes the image enhancement algorithm based on improved Dark Channel Prior algorithm to preprocess on monocular vision image. Firstly, the paper build degradation model of underwater image color slants and atomization phenomenon. Then, the image of depth information is calculated through the parallax of brightness channel and dark channel to estimate the background color of the water accurately. On this base, the adaptive scaling factor selection strategy is adopted to process transmission diagram for the higher contrast image restoration effect. In addition, the color correction method is taken to remove residual color slants and enhance the overall brightness of the image. The experiment result shows that the presented method can have better resolution and brightness of visual effect than the conventional one.