一种有效的图像去雾方法

摘要

使用一种有效的方法实现雾化图像的去雾处理.首先根据雾天图像退化的物理模型,基于暗原色先验的单一图像去雾算法原理实现图像去雾的粗处理;然后针对粗处理后的图像,提出一种简单的去块滤波方法,有效地消除块效应;最后将图像转换到HSV颜色模型下,对亮度进行处理,降低色彩失真.实验结果表明,我们所提出的方法能够显著地改善有雾图像的视觉效果.%In this paper, we propose a method which can effectively remove fog effect from images. Based on the model of foggy image degrading and the principle of dark channel priority, foggy images are processed roughly. Then, a simple deblocking filter is used to remove the blocking effects. Finally, under HSV color space, we adopt nonlinear transformation of brightness to reduce the distortion in color. The algorithm has been evaluated on images with different foggy concentration under different scenes. Experimental results show that the proposed approach can improve the visual effect of image well.