Abstract: Haze is a physical phenomenon that considerably degrades the visibility of outside scenes. This is often in the main because of the atmosphere particles that absorb and scatter the sunshine. This paper introduces a unique single image approach that enhances the visibility of such degraded pictures. Our methodology may be a fusion-based strategy that derives from 2 original hazy image inputs by applying a white balance and a distinction enhancing procedure. To mix effectively the knowledge of the derived inputs to preserve the regions with sensible visibility, we tend to filter their vital options by computing 3 measures (weight maps): brightness level, colour property, and strikingness. To reduce artefacts introduced by the burden maps, our approach is intended in an exceedingly multiscale fashion, employing a Laplacian pyramid illustration. We tend to square measure the primary to demonstrate the utility and effectiveness of a fusion-based technique for dehazing supported one degraded image. the strategy performs in an exceedingly per-pixel fashion, that is easy to implement. The experimental results demonstrate that the strategy yields results comparative to and even higher than the additional complicated progressive techniques, having the advantage of being applicable for period applications.
Keywords: Dehazing, Strikingness,Fusion based, single image, Weight maps.