Abstract: In view of fuzzy and degenerated images in coal mine environment due to presence of large amounts of coal dust and water mist, a sharpening algorithm based on dark primary principle and principal component analysis was proposed. Based on atmospheric scattering model, transmittance is calculated according to the dark primary principle. The principal component analysis is used to obtain brightness, saturation and contrast, which can fully reflect fog image information. Atmospheric light value is calculated by weighting these indexes, so as to realize sharpening process of underground images with fog and dust in underground coal mine. The simulation results show that the proposed algorithm can restrain image detail to a great extent, maintain authenticity and structural integrity of the image, and have good real-time performance.