The imaging process of remote sensing images is easily affected by fog. At the same time, the increase of foggy weather makes the fog removal work of remote sensing images become an important part of remote sensing image data preprocessing. The dark channel fog removal method based on dark color statistical priority is widely used in the defogging of natural images. But for remote sensing images, the applicability of the dark channel fog removal method needs to be evaluated. Taking "GF-1" satellite remote sensing images as an example, the dark color characteristics of remote sensing images are discussed. The result shows that the dark channel fog removal method based on dark color statistic priority can be applied to the fog removal of remote sensing images.