| In aerographily, the visibility is the most distancy that an object can be identified clearly. it is related with the weather of the day closely. The contrast and the color of the scenery which is screened in the low visibility (such as fog, haze, dust, dusk and night) will be changed with the influence of the weather, many characteristics of the image are covered up, and then the identifiable degree greatly debased. This situation will bring serious effect to the quality of the video image of the system of the traffic, outdoor inspector, spy, navigation and so on. So recovered and enhanced the scenery's detail information from the degenerated image will have great realistic meaning.There are two different kinds of ways to make the image clear, that is image enhanced and image recovered. Image enhanced takes no account of the reason of the degeneration and only according to the specific need extrude or restrain some information of the image. Image recovered checks the reason of the degeneration, including the image recovered based on degeneration model and blind recovered based on no transcendental knowledge. In this paper, the major task is to do the research in the algorithm of the enhanced and the recovered.Firstly, the paper does some research in the celerity by using the algorithm of the enhanced aiming at the low visibility image, and brings forward a Retinex Enhanced Image Enhanced Algorithm Based on Color Constancy, after enhanced the low visibility image of different weather, and get a clearer image. The result of the experiment shows that this algorithm can enhanced the discrimination of the scenery and the contrast of the image. Secondly, aiming at the research of the algorithm of the image recovered, bring forward a new algorithm of blind separated recovered, after blind recovered the low visibility image, and get a clearer image, the result of the experiment shows that this algorithm can recover the image's detail character without regard to the degenerated model, and have the better recovered effect. |