Research Of Fog-degraded Image Restoration Techniques

Because of atmosphere scattering, images which are captured under bad weather conditions such as fog, mist, rain, and snow have poor contrasts and colors. Current outdoor vision algorithms only work effectively in clear weather. So it causes problem to the outdoor visual system. In order to develop outdoor visual system that perform well under all weather conditions, it is essential to model the visual effects of various weather conditions and develop algorithms to remove them.The restoration of fog-degraded images is a very challenging problem. Recently, the research of fog-degraded image restoration is still on the stage of starting. There has been an increased interest on this problem. We concentrated image restoration techniques based on atmospheric scattering model in this paper.The work in this paper firstly analyze the common methods of image enhancement such as histogram equalization, local histogram equalization, interpolating histogram equalization and their applications on fog images. Unfortunately, it is well known that the effects of bad weather increase exponentially with the distance of scene points from the sensor. As a result, conventional techniques based on space invariant filtering can not remove weather effects from degraded images completely. To solve the problem of restoration of fog-degraded images, we address image restoration from two images of the same scene under different weather conditions using simple additional information provided interactively by the user. The methods based on inverse-intensity chromaticity space were also introduced and applied to the physics-based models. After that, some enhanced images with good visibility were demonstrated. Furthermore, our interactive method of image restoration can be used for a variety of image processing software.