Single image haze removal using integrated dark and bright channel prior

In recent years, the dark channel prior (DCP) has been proven to be an adequate haze removal model. However, its procedure causes annoying halo and gradient reversal artifacts. To remove these issues, numerous filtering techniques have been designed and integrated with DCP. But, these filters inevitably induce massive computational load and edge degradation problems. In this paper, a novel visibility restoration model is proposed to solve the above-mentioned problems. It utilizes DCP, bright channel prior (BCP) and gain intervention filter. BCP is used to solve the sky-region problem associated with DCP based dehazing. The gain intervention filter is used to improve computational speed and edge preservation. The experimental results reveal that the proposed dehazing model provides better computational time as compared to the existing dehazing techniques. It also provides superior restored images over existing dehazing techniques. More significantly, the designed dehazing model possesses highest potential for real time applications due to its superior restoration results and computational speed.