针对经典的暗原色理论算法在处理雾天图像时，图像的边缘光晕现象，图像的色度、色调、亮度失真等问题，提出基于全变差(TV)模型的暗原色理论雾天图像增强算法。基于暗原色理论数学模型，粗略的获取透射率图像，并在此基础上，通过引入TV 模型，对透射率图像进行保边平滑优化，为避免图像中高亮度区域出现光晕失真现象，结合容差机制，判断图像中的天空区域，对透射率图像进行修正，根据暗原色理论的数学模型反向求解，获得复原后的图像。通过主观观察和客观评价，在整体和细节方面该算法比经典的暗原色算法有更好的处理效果。

To deal with the hale phenomenon at the image edges, image hue、tone and brightness distortion problems in the classic dark channel theory algorithm, enhancement dark channel algorithm of fog image based on the total variation (TV) model is proposed.The rough transmission image based on the dark channel theory mathematical model is smoothed by the TV model. The image is segmented by the tolerance mechanism and distinguishes sky areas, so as to solve the halo phenomenon in the image highlight area. According to the dark channel theory mathematical model, it gets a defogging enhancement image from the inverse solution. Through the subjective observation and objective evaluation, the algorithm is better than the classic dark channel algorithm in the overall and details.