To improve the quality of the sea surface image under foggy weather, we propose an innovative single-image defogging algorithm based on the dark channel prior principle. The median filter is combined with the minimum filter for the purpose of obtaining accurate dark channel values in areas where the depth of field changes sharply. Then, the algorithm constructs a fog density detection model to obtain the atmospheric light value from the dense fog area and effectively eliminate the interference of bright spots. Subsequently, the atmospheric scattering model is applied to obtain a preliminary fog-free image. Finally, an adaptive logarithmic mapping algorithm is introduced to enhance the visual effect of the defogged image. Experimental results show that the proposed algorithm can effectively improve image quality degradation and avoid halos in the intersection region of sky and sea, Moreover, the method does not require guided filtering for transmittance refinement, which greatly improves the execution speed of the algorithm.