基于暗通道的单幅图像快速去雾算法

为了提高雾天环境下退化图像的视觉效果,以大气退化模型和暗通道先验估计为基础,提出了一种快速单幅图像去雾算法.利用暗通道先验对大气透射率进行粗估计,用采样边缘保持滤波细化透射率,再设法估计出原图中的天空区域,对其进行透射率补偿,根据估计的大气光和透射率,利用大气退化模型即可恢复出无雾图像.针对去雾后的图像色度较暗的问题,使用一种自适应直方图均衡方法.实验结果表明,相对于已有算法,该算法有效提高了图像去雾的效果,而且加快了处理速度.

Fast defogging algorithm for single image based on dark channel

In order to improve the visual effect of degraded images in foggy environment,a fast dehazing algorithm for single image is proposed based on the degradation model and the dark channel prior estimate.Using dark channel prior estimates the atmospheric transmittance roughly,then using edge preserving filtering estimates the refinement of transmittance,after that try to estimate the sky region in the original image for compensating of the transmission of the region.Finally,according to the estimated atmospheric light and transmissivity,the fog image can be recovered by the atmospheric degradation model.In order to solve the problem that the color of the image is darker,the adaptive histogram equalization method is used.The experimental results show that compared with the existing algorithms,the proposed algorithm can effectively improve the effect of image fogging and speed up the processing speed.