基于暗通道先验的快速图像去雾算法

探讨了暗通道先验去雾算法的原理,针对暗通道先验去雾算法时间复杂度太大的缺点,提出用快速有效的巴特沃兹低通滤波器代替复杂的软抠图方法实现对透射率的平滑与细化;针对暗原色图像在景深交界处存在白边现象采用求区域最大值法加以修正;并给出了自适应的求解全局大气光算法.实验结果表明,改进的暗通道去雾算法在获得满意的图像去雾效果的同时能大大提高图像去雾算法的速度,能满足工程上的实时应用要求.

Fast Image Dehazing Method Based on Dark Channel Prior

The principle of dark channel prior to fog algorithm was discussed in the paper,and the soft matting smooth and refined transmittance was achieved using a quick and efficient Butterworth low-pass filter instead of the complex method according to the shortcoming of time complexity of the algorithm. For the white edge image existing at the junction of the phenomenon in depth of dark colors,the method of seeking the maximum area was used,and an adaptive algorithm for solving global atmospheric light was also given in the paper. Experimental results showed that the improved algorithm in the dark passage to fog the image not only obtained satisfactory results but also greatly increased the speed of the image to fog algorithms to achieve the real-time applications requirements of engineering.