Images and videos taken outside the foggy day are serious degraded. In order to restore degraded image taken in foggy day and overcome traditional Dark Channel prior algorithms problems of remnant fog in edge, we propose a new dehazing method.We first find the fog area in the dark primary color map to obtain the estimated value of the transmittance using quadratic tree. Then we regard the gray-scale image after guided filtering as atmospheric light map and remove haze based on it. Box processing and image down sampling technology are also used to improve the processing speed. Finally, the atmospheric light scattering model is used to restore the image. A plenty of experiments show that algorithm is effective, efficient and has a wide range of application.
Real-time single image dehazing based on dark channel prior theory and guided filtering
AOPC 2017: Space Optics and Earth Imaging and Space Navigation ; 2017 ; Beijing,China
Proc. SPIE ; 10463
2017-10-24
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Real-time single image dehazing based on dark channel prior theory and guided filtering [10463-59]
British Library Conference Proceedings | 2017
Prior based Single Image Dehazing using Decision Image
IEEE | 2020
|Springer Verlag | 2021
|