Abstract: Aiming at the problems of environmental light color radiation and color degradation caused by age in the shooting process of mural images, a color restoration and enhancement algorithm of Dunhuang murals under the fusion of multi-task decomposition and adaptive color equalization is proposed for the first time to solve the above problems. By analyzing the problems existing in the mural image, the dark channel information is first extracted from each channel of the mural and the texture and detail are enhanced. At the same time, in order to avoid the dirty points on the original mural after enhancement affecting the color enhancement effect, the local filtering processing is carried out, and the adaptive color equalization model is used to make the output mural not only improve the color, but also enrich the texture information. Then, the fusion algorithm is used to fuse the results of adaptive color equalization and filtering, which makes the details and color information of the mural more abundant, and the contrast, energy value and correlation of the mural image are greatly improved.