基于暗原色和加权形态学滤波的图像去雾算法

针对雾天图像能见度低、对比度差的特点,提出一种自动消除雾的方法：基于暗原色和加权形态滤波的增强算法。首先引入暗原色先验信息,然后利用形态学滤波方法估计雾浓度图。该方法既能平滑雾浓度图,又能很好地保留场景的边缘,使估计出的雾浓度图更加精确。最后恢复去雾图像。实验结果表明,该方法简单快速有效,能够很好地达到去雾目的,并且较好地保留图像边缘细节。

Auto image haze removal algorithm based on dark channel prior and weighted morphological filtering

According to the low visibility and poor contrast of haze images,a new automatic enhancement algorithm based on dark channel prior and weighted morphological filter is proposed.The information of haze density is obtained with dark channel prior.The haze density is estimated with morphological filtering method which can smooth the haze density map and preserve edges of scene,and may make the estimated haze density map more accurately.Finally,a high quality haze-free image is recovered.The experimental results show the proposed method is simple,fast and effective for obtaining good results,and keep the detail edges of images.