Scene Priors-Based Restoration Method for Heavy Fog Airport Runway Images

Restoring a clear runway scene from hazy and foggy images is crucial for visual navigation. However, some heavy fog scenes cannot be restored by using conventional defogging methods. Instead of only depending on defogging methods, this paper proposes a novel approach which combines scene prior images with defogging algorithms to recover heavy fog runway images, which is designed as the SPBR method. Meanwhile, a brand-new defogging network ReAOD-Net is designed to achieve better defogging visual performance in our algorithm based on the fundamental principles of AOD-Net. The experimental results demonstrate that the proposed method can better restore the fog-free scene in heavy fog images with prior information available compared with the existing defogging methods.