Single Image Dehazing Using Manifold Particle Filter
Bingyuan Wang, Zheng Fang, Xia Li, ShuaiHua Zhang · 2019
Degraded images taken in bad weather such as fog or haze puts negative impacts on automation without pre-processing. To eliminate halo artifacts and improve the image visibility, in this paper, based on the dehazing algorithm of the dark channel prior, a manifold particle filtering algorithm is proposed. Specifically, the proposed technique uses manifold particle filter to optimize the transmission and achieve accurate transmission estimation. The proposed algorithm was compared with existing defogging techniques on nature foggy images and synthetic images. Experimental results demonstrate that the proposed algorithm is capable of removing halo artifacts and restoring the visibility of foggy image, which also achieves higher restoration quality, both quantitatively and qualitatively.