Abstract: In order to take full advantage of the relationship between reference color images and the objective grayscale image and to improve the degree of automation for image color reconstruction,we presented an automatic algorithm for image colorization based on dictionary learning and sparse representation.Firstly,a joint dictionary is trained by reference color images according to the correlations among the luminance,feature and color of trained images.And then the sparse coefficients under the joint dictionary for the objective grayscale image are computed by using its luminance and feature information.Finally,the color information is reconstructed using the above joint dictionary and the obtained sparse coefficients.Image segmentation is not necessary in the proposed algorithm.The color reconstruction is made on the entire image and therefore the proposed algorithm is global and automatic.Experimental results demonstrate that the algorithm presented in this paper is effective and efficient,especially for those monotonous images.