A novel image defogging algorithm based on multi-resolution fusion transform

An emerging trend in the field of image restoration is the removal of haze or fog from an image or video sequence to improve the quality of the image. Such image restoration techniques is widely used in applications like traffic monitoring and surveillance during hazy weather conditions, prediction and analysis of volcanic activities, etc. In this paper, we propose a novel algorithm based on a fusion model integrated with a multi-resolution approximation technique. The technique decomposes the given hazy image into its frequency components in which the most distinct feature values are extracted using a fusion model. Our proposed algorithm is tested with various hazy images under varying degrees of fog. Experimental results show that the proposed approach is efficient and efficient for foreground object detection and visibility enhancement under fog weather conditions.