Image Haze Detection and Removal Based on Multiresolution Spatial Analysis

In this paper,two procedures have been developed for haze removal using multiresolution spatial analysis. The image contaminated by haze is decomposed into layers of different resolutions of spatial distribution frequencies. Although haze is distributed in the lower frequency layer, this layer may also contain a component of land cover that is spatially and temporally relatively stable. For multitemporal reference image correction method, a haze-free reference image of the same area is used to characterize land cover with multiresolution spatial analysis. The component of the characterized land cover is removed from the overall low frequency content of an image,the remaining haze contribution can be subtracted from the image to obtain an estimated haze-free rendition. For fully automated haze removal method, the radiometric characteristics of the corresponding layers are estimated and analyzed with topology-based multiresolution spatial analysis technology. Based on the feature and texture analysis, the haze component is then separated from the remaining spatial frequency components representing spectral information of actual land cover types in the scene, and a spectrally corrected image with "haze-off" characteristics is obtained. Case study using several different type images of Qingdao City in China proves the effectiveness of these techniques except for those regions too hazy.