Image-Dehazing-based-on-ROP

Image dehazing was a type of scene recovery designed to reduce or eliminate the effects of image blurring and reduced contrast caused by haze, atmospheric pollution, or other atmospheric particles. In this article, a real-time scene recovery framework that could be used to recover degraded images under different environments, including underwater, sandstorm, and haze, was described. Degraded images could actually be considered as a superimposition of a clear image with the same color imaging environment (underwater, sandstorm, haze, etc.). In order to characterize this phenomenon mathematically, rank-one prior (ROP) could be introduced. In this article, six new frameworks of dehazing algorithms were developed by combining ROP with Color Transfer and six image sharpening techniques, AMF-UM, AMF-NS, GaussianF- UM, GaussianF-NS, GuidedF-UM and GuidedF-NS. Comprehensive experiments on scene recovery showed that the six methods proposed in this article outperformed ROP+ in terms of efficiency and demonstrated a very high degree of superiority in terms of performance.