A novel Edge-Aware Weighted Guided Image Filtering

Abstract

Halo artifact is the main problem in existing techniques for filtering based image preserving as well smoothing techniques. Halo artifact is nothing but the uneven lightning conditions through the image. To solve the problems of existing techniques we implemented a novel weighted guided image filter (WGIF) which is integration of two techniques one is guided image filtering and other is edge aware weighting. The proposed integrated method is having two main advantages, one is the run time complexity of proposed WGIM is same as the GIF and the other advantage is the halo-artifacts present in existing technique is completely removed by this technique. The proposed work may be used for different applications like removal of haze from single image, image fusion for different exposed image as well as single image detailed enhancement.Matlab execution results will show that the proposed work is having ability to provide the best results for local and global smoothing filter advantages as well as the halo artifacts present due to uneven distribution of light is successfully removed from the image. We extended this implementation for videos as we know the video is nothing but number of frames. So the same implantation is applied for video and quality analysis is done.