Litcius/Paper detail

Saturation Based Iterative Approach for Single Image Dehazing

Zongwei Lu, Bangyuan Long, Shiqi Yang

2020IEEE Signal Processing Letters38 citationsDOI

Abstract

In this letter, we explore the dark channel prior from a new perspective. By re-formulating the dark channel, it is found that the dark channel correlates closely with the saturation and brightness. A novel way to estimate the transmission without the need to compute the dark channel is then presented and it can prevent the transmission from being under-estimated. In order to discourage the scene radiance from becoming over-saturation and remove the haze effectively, an iterative approach with a tolerable bound on the saturation is proposed. Qualitative and quantitative experimental results demonstrate that the proposed algorithm can effectively restore the scene radiance with a comparable or better visual quality than many other algorithms. Furthermore, it runs faster than most state-of-the-arts.

Topics & Concepts

RadianceComputer scienceSaturation (graph theory)BrightnessChannel (broadcasting)Artificial intelligenceComputer visionIterative methodAlgorithmOpticsPhysicsMathematicsTelecommunicationsCombinatoricsImage Enhancement TechniquesAdvanced Image Processing TechniquesAdvanced Vision and Imaging
Saturation Based Iterative Approach for Single Image Dehazing | Litcius