Image Defogging algorithm Based on Image Bright and Dark Channels

This paper proposes an image defogging algorithm based on the combination of bright and dark channel in fog and haze weather. A model of the air light scattering is proposed based on the physical model of degraded image. Air light value and transmissivity are estimated by using the combination of light channel prior and dark channel prior. The algorithm can solve color distortion problem of sky area when fog-free image is restored, recover the image details and color, and improve the vision effect of the image. Evaluation parameters are used to compare the image quality. Simulation results show that the algorithm proposed in this paper is better than multi-scale Retinex image defogging algorithm.