Color Transfer Between Images By Minimizing Corruptive Artifacts

Color transfer between images is a critical operation in image editing but easily suffers from some corruptive artifacts in the mapping process. In this paper, a color transfer approach with corruptive artifacts suppression is define, which performs iterative probabilistic color mapping. It is done with the help of self-learning filtering scheme and multiscale detail manipulation scheme, which minimizes Kullback-Leibler distance. First, an iterative probabilistic color mapping is applied to construct the mapping relationship between the reference and target images. Then, a self learning filtering scheme is applied into the transfer process to prevent from artifacts and extract details. The transferred output and the extracted multi-levels details are integrated. This is done by the measurement minimization to yield the final result. This method achieves a sound grain suppression, color fidelity (the degree to which the output image matches the original images) and detail appearance.