Visible light imaging of target scene under the condition of rain and fog plays an important part in the simulation of battlefield environment. It is of great significance to enhance visible image to improve the range and precision of the optical guided missile. In an experimental cabin under atmospheric environment, the simulation of target mechanical structure named miniature landscape sand table is carried out under various rain and fog conditions to obtain visible images in sunny day, mist, fog, dense fog, light rain, moderate rain and heavy rain. The visible images are enhanced based on the improved dark channel method, and, the mean squared error (MSE) and peak signal to noise ratio (PNSR) of the rain and fog images and the enhanced images are calculated. The calculated results show that , with the increase of rain and fog level, MSEs of the rain and fog images and enhanced images gradually increas from 245.78 to 383.07, and from 46.84 to 222.01, respectively, and the PNSRs gradually decrease from 24.23 to 22.29, and from 31.42 to 24.66, respectively. The quality of the images decreases with the increase of rain and fog level, even after the images are enhanced. In addition, compared MSE and PNSR of the rain and fog images with those of the enhanced images, it shows that MSE decreases obviously and PNSR increases significantly. The quality of the images is improved significantly according to the two objective indexes of MSE and PNSR.