Atmospheric light estimation using fog line vector for efficient defogging without color distortion

Abstract

This paper presents an efficient defogging method with less color distortion. The proposed method consists of three steps: i) estimate atmospheric light using fog line vectors and fog-free reference image; ii) estimate transmission using fuzzy membership function; iii) refine the transmission using L1-Norm regularization, and obtain a defogged image based on fog model using the estimated the atmospheric light and the refined transmission. Experimental results show that the proposed method can effective remove fog components in the foggy image without color distortion.