Weighted Guided Image Filter for Smoothing and Haze Removal

Abstract

Most applications in computer vision and computer graphics involve image filtering to suppress and extract content in images. In this applications Edge preserving smoothing is the concerned area. There are different edge preserving filters. Those are global filters and local filters. Local filters like guided filter is very simple but it produces halo arifacts. Global filters like Weighted least square filter preserve sharp edges but have high computational complexity. So we proposed Weighted guided filter by incorporating edge aware weighing into the guided filter. It preserves sharp edges like global filters and it is very simple as local filters. Due to its simplicity it is used in different applications like single image haze removal. Remote sensing images such as satellite and underwater images are widely used in various fields of computer vision. But due to fog, mist and various aerosals in the atmosphere their contrast get reduced. So here for single image haze removal  weighted guided image filter is used.