Haze removal using visible and infrared image fusion
Byungtae Ahn, Tae-Wuk Bae, In-So Kweon · 2011
In order to remove haze from images, we propose an improved method for fusing a visible and a near-infrared (NIR) image of the same scene. MR images are inherently haze-free. The NIR information allows dehazing without building and analyzing a model of haze image. We suggest a multi-resolution fusion using edge-preserving filter, bilateral filter, to reduce artifacts and computational time compared to the existing methods. Experimental results validate our approach.