“Color Transfer” by Pouli and Reinhard

  • ©Tania Pouli and Erik Reinhard



Entry Number: 27


    Color Transfer

Course Organizer(s):




    Intended Audience
    Anyone who creates or edits 2D content, including technical directors, artists, photographers, game designers, and filmmakers. Also image-processing researchers who may discover opportunities to develop further techniques or enhance their own algorithms.

    Color transfer was introduced about 10 years ago as a technique to make rendered images look more natural by adjusting their color content on the basis of an example image. Now, color transfer is not a single algorithm but a range of methods and techniques that aim to make one image look more like another, and the technique now has applications far beyond its humble beginnings.

    This course helps color-transfer researchers understand where there may be further opportunities for algorithmic improvements, and it helps practitioners in creative industries such as photography, movies, and games understand how to make the most of these algorithms. The course begins with an overview of current techniques, and then presents many examples and comparisons that show when these algorithms are expected to produce their best results in still images and videos. It also shows how to choose appropriate examples to steer the results and reviews applications of color transfer that include making night-time images from day-time images, color correcting stereo pairs, and color matching photographs to stitch pre-processing to panoramas.


Overview Page: