For Full-Text PDF, please login, if you are a member of IEICE,|
or go to Pay Per View on menu list, if you are a nonmember of IEICE.
Color Improvement Based on Local Color Transfer in Digital Intermediate of Digital Cinema
Ho-Gun HA Dae-Chul KIM Wang-Jun KYUNG Yeong-Ho HA
IEICE TRANSACTIONS on Information and Systems
Publication Date: 2015/01/01
Online ISSN: 1745-1361
Type of Manuscript: PAPER
Category: Image Processing and Video Processing
digital cinema, color transfer, recoloring, image filtering,
Full Text: PDF(4.7MB)
>>Buy this Article
In digital cinema, an image goes through many types of processes like scanning, mastering, and digital intermediate. Among them, the digital intermediate process plays a central role because it determines the final color of an image. It edits and changes the colors of the images. However, some color distortions such as color bleeding are generated when editing and changing local colors in an image. In this paper, local color improvement for digital intermediate is proposed based on color transfer. Our method is simple and efficient color improvement that does not requires neither precise image segmentation nor feature matching. To prevent color distortions, a modified color influence map is proposed with color categories. First, the source image is roughly segmented using a color category map, which groups similar colors in color space. Second, the color influence map is modified by assigning different weights to the lightness and chroma components. Lastly, the modified color influence map and color category map filtered with anisotropic diffusion are combined. Experimental results show that the proposed method produces less color distortion in the resulting image.