HAZE REMOVAL AND DETAIL ENHANCEMNT ALGORITHM USING GLOBALLY GUIDED FILTERING

Abstract

Nowadays, the utilization and demand of image filters is very high in various applications of the areas of computer vision and image processing. Many local edge preserving smoothing strategies including Guided Filter [2] and Weighted Guided Filter [3] have been proposed in the literature earlier. These have been applied for studying different applications such as detail enhancement, de-hazing, feathering and many more. But these techniques may fail to preserve fine structured regions. A new technique, i.e. globally guided filtering has been introduced for solving this issue.

In this study, I utilized the functionality of globally guided filter [1] in the de-hazing application as well as for the detail enhancement of the images. We have proposed a model for detail enhancement using globally guided filtering by incorporating edge-based weighting in it. Initially, a model was designed which was based on decomposition algorithm for detail enhancement. Then, the results are obtained for both of the algorithm models and compared. Whereas this model is having some limitation including noise amplification in flat patch areas and halo artifacts on edges. Therefore, edge-based weighting is incorporated to resolve this issue.