The human visual system processes lightness constancy to estimate reflectance under complex viewing conditions. We focus on the underlying reason the human visual system is conscious of haze, and discuss the single-image haze-removal problem based on lightness constancy from the beginning of haze illusion. Following the basic principle and model of "atmospheric transfer function", we bring about a simple dehazing paradigm by estimating two values using bilateral filtering. Practical investigation of parameters and design are analyzed in detail. Furthermore, a novel quantitative standard haziness presents itself naturally during the estimation process. Comparative study and objective evaluation demonstrate that the proposed method is fast and effective, yielding high-contrast and vivid haze-free images.
© 2012 Chinese Optics Letters
Cheng Li, Jianzhong Wang, Qingshun Han, and Duyan Bi, "Lightness constancy: from haze illusion to haze removal," Chin. Opt. Lett. 10, 081001- (2012)