This work presents an effective method to colorize the Chinese ink-and-wash paintings. The Chinese ink-and-wash paintings are first digitized into the images. The user can impose the initial colors on the image freely to become the segmented-marked image. Using image sub-sampling to downscale the segmented-marked image as the sub-image, the image colorization is then implemented on the sub-image. Since the number of image sub-sampling process is very flexible, the proper number will efficiently accelerate the colorization procedure and take fewer computing time. In addition, the colorization procedure is utilized to refine the boundary between the different colors in the up-scaled colorized subimage. The lowpass filter is applied on the last colorized image in order to present the soft-gradual tone in the ink-and-wash paintings, such as, water-flowing, smog, cloud, waterfall, and shadow etc. Simulation results have compared the image quality and the computing time for Levin’s and the proposed methods. Both two methods result in similar quality colorized images. However without the cost function involved, the proposed method needs 1/15 to 2/3 times the computing time of Levin’s method.
展开▼