Fast algorithm for restoration of foggy images
Artur Łoza, Harish Bhaskar, Mohammed Ebrahim Al-Mualla, David Bull · 2013
This paper describes a new fast method for foggy images restoration derived from the dark channel prior methodology. Based on a physical foggy image formation model, the parameters of the restoration algorithm are estimated from the saturation channel of a single colour image without any prior knowledge of the scene. The performance of the proposed method is evaluated on a set of real-world images and compared with other state-of-art fog removal and contrast enhancement methods. It is shown that the new method outperforms other tested techniques in terms of the execution speed and perceptual quality of the restored images.