Haze Removal: An Approach Based on Saturation Component

Abstract

AbstractOutdoor images those are taken under bad weather conditions are basically degraded by the various atmospheric particles such as smoke, fog, and haze. Due to the atmospheric absorption and scattering phenomena while capturing the images, the irradiance received by the camera from the scene point is attenuated along the line of sight. The incoming light flux is attenuated with the light from all other directions called the airlight. Due to this reason, there is a resultant decay in the color and the contrast of the captured image. Haze removal from an input image or dehazing of an image is highly required so as to increase the visibility of the input image. Removing the haze layer from the input hazy image can significantly increase the visibility of the scene. The haze-free image is basically visually pleasing in nature. The paper focuses on the haze removal process by considering the HSI color model of an image instead of RGB color space. In the HSI color model, the saturation component describes the contrast of an image. From the saturation component, it is possible to estimate the transmission coefficient or the alpha map, and from this, a haze-free image can be recovered which has the better visibility than that of the captured hazy image.KeywordsScatteringAirlightAttenuationHazeSaturationImage modelingTransmission coefficient