Murals are the important components of culture and arts of Dunhuang. Unhappily, these murals have been ruined or are being ruined by some diseases such as cracking, hollowing, falling off, getting mildewed, dirt, and so on. Due to a lack of a standard mural datasets, Dunhuang mural datasets are created by ourselves. Meanwhile, our proposed network architecture SeparaFill which is connected two generators based on U-Net. First, the contour restoration generator network is used to repair contour lines. Then, the color mural image is repaired by the content completion network with help of the repaired contour. Next, global and local discriminant networks are applied to determine whether the repaired mural image is authentic in terms of both the modified and unmodified areas. Compared with existing mural restoration algorithms, the proposed method increases the peak signal-to-noise ratio (PNSR) and increases the structural similarity (SSIM). SeparaFill shows good performance in restoring the line structure of the damaged mural images and retaining the contour information of the mural images.