基于暗通道先验的全局去雾算法

Global Defogging Algorithm Based on Dark Channel Prior

针对雾天采集到的图像存在清晰度低、对比度低等问题,提出一种基于暗通道先验的全局去雾算法。该算法首先获取图像RGB整体均值与各个通道之间的均值关系,以及RGB通道中每个像素点与其相对应通道光照均值之间的关系,结合大气散射模型进行恢复。实验表明,该算法进行在去雾效果上提升了图像的对比度和清晰度,并提高了图像的细节信息。

For addressing the problems in the images collected in the foggy whether,such as low resolution and poor contrast,a global defogging algorithm is put forward based on dark channel prior.The algorithm obtains the relationship between the overall RGB mean value of the image and the mean value of each channel,as well as the relationship between each pixel in the RGB channel and the mean value of illumination in its corresponding channel,and then restores the image with the atmospheric scattering model.Experiments show that this algorithm can improve the contrast and clarity of the image and improve the details of the image.