Abstract: Due to the poor underground environment, there is a lot of dust and mist, which makes the images of underground video in coal mines seriously degraded. Concerning that the existing transcendental dust and mist clearness algorithm based on dark channel has limitation in dealing with dust and mist degraded images, we propose a kind of restoration algorithm for dust and mist degraded images based on improved negative correction. For the serious halo effect of original algorithm, we establish the mapping among the parameters to achieve the correction parameters refinement. Considering that the brightness of restored images is relatively low, Gama correction is adopted to get final clear images. The experiment results show that the proposed algorithm can effectively achieve the restoration of dust and mist degraded images, and the restored images have more saturated colors, richer information compared with other algorithms, which demonstrates the superiority of the proposed algorithm.