An image defogging enhancement algorithm combining dark channel and haze-line prior

Abstract

Affected by dust and haze weather, outdoor images will appear distortion and fuzzy to a certain extent. Aiming at the characteristics of dark channel prior and haze-line prior, an improved fog removal method combining the two methods is proposed, which can effectively restore the fogging image of the sky region. Firstly, the image is divided into sky region and non-sky region by threshold segmentation, and the airlight value is estimated by introducing brightness and gradient information into the image. Secondly, dark channel prior model and haze-line prior model are used to estimate the transmission of foreground region and sky region respectively, then gamma correction and guided filtering are used to optimize the transmission map of the whole image. Finally, combined with the hazy image formation model, the restored image is obtained by inverse operation. Then the brightness compensation and contrast stretch correction are used to restore the brightness and contrast of the image, and the final defogging image is obtained. Experimental results show that this algorithm can effectively solve the image distortion, improve the contrast and sharpness, so as to get a higher quality image.