Performance evaluation of various desmogging techniques for single smoggy images

Abstract

The worsening air pollution is producing serious smog problems in the world. The high concentration of aerosol, that attenuates the scene radiance and adds undesired scattering illumination into the actual illumination values, has a serious effect on the visibility of the images. Therefore, images captured under smoggy environments suffer from poor visibility. Because, the smog particles attenuate the illumination reflected by the targets and add undesired scattering light. Therefore, imaging under smoggy environments affect the performance of many machine vision systems such as intelligent transportation system, remote sensing imaging, aerial imaging, etc. From the literature, it has been observed that majority of existing researchers have either focused on foggy or hazy images only.