考虑亮通道取反的修正暗原色先验去雾算法

Image dehaze algorithm considering inversed-bright-channel to redress dark channel prior

为提升图片质量,改善传统的暗原色先验处理图像的局限性,提出一种RGB亮通道取反的自适应修正暗原色先验去雾算法。在计算暗通道前,先对有雾图像的RGB通道中最亮通道进行取反处理,自适应修正透射率的计算值;在全局大气光估计时,采用最亮0.1%的像素的均值;使用canny算子提取图像的边缘信息,对导向滤波的方法进行修正,进行透射率估计;根据大气散射模型对有雾图像进行复原。实验结果表明,该算法可以有效改善暗原色先验在天空区域失效的问题,能够保留图片的景深特征,使复原图像具有更清晰的边缘细节。

To improve the image quality and overcome the limitations of traditional dark prior processing,a defogging algorithm by adaptively rectifying dark channel prior with inversed RGB bright channel was proposed.Before calculating the dark channel,the brightest channel among the RGB channels of the haze image was inverted,and the calculated value of the dark channel was adaptively redressed.The mean value of the top 0.1 percent brightest pixels was used in the global atmospheric light estimation.Canny operator was used to extract the edge information of the image,and the guided filtering method was modified to estimate the transmittance.The haze image was restored according to the atmospheric scattering model.Experimental results show that the proposed algorithm can effectively improve the problem of dark prior failure in the sky region?preserve the depth of field features of the image,and make the restored image have clearer edge details.