Single-Image-Based Rain and Snow Removal Using Multi-guided Filter

Abstract

In this paper, we propose a new rain and snow removal method through using low frequency part of a single image. It is based on a key difference between clear background edges and rain streaks or snowflakes, low frequency part can obviously distinguish the different properties of them. Low frequency part is the non-rain or non-snow component. We modify it as a guidance image, the high frequency part as input image of guided filter, so we get a non-rain or non-snow component of high frequency part and add the low frequency part is the restored image. We further make it more clear based on the properties of clear background edges. Our results show that it has good performance in rain removal and snow removal.