De-Hazing using Guided and L0 Gradient Minimization Filters

Objectives: Images captured under foggy or hazy conditions have low contrast and meager color. The proposed work is carried to improve the visibility of hazy images. Methods: The proposed method is based on ‘dark channel prior’ to remove haze, along with guided filter to refine the image due to the occurrence of halos. Even after refinement, halos are not completely removed. So, for further refinement edge preserving smoothing global filter L0 gradient minimization filter is used. Findings: This method has an advantage over conventional method by suppressing halos completely. Application/Improvements: The proposed method improves the contrast and increase in entropy reveals that the detail information is preserved effectively.