Title :
Color Transfer into Scenes
Author_Institution :
State Key Lab. of Comput. Sci., Inst. of Software, Beijing, China
Abstract :
In this paper we present novel techniques to redesign materials of interior virtual scenes. Users would need to provide no more than one example image which has desirable color distribution. Our system modifies the materials of the scene so that the rendering result would similar to the color style of the example image. We first extend ISODATA clustering algorithm to retrieve representative viewpoints from a large amount of viewpoints distributed in the scene. Then, around each clustering center, an environment map image will be generated to embody global illumination radiance of the scene. Finally, by recoloring the environment map image set, the materials of objects will be redesigned accordingly. Through several iterations, the user can design aesthetically pleasing materials as their will.
Keywords :
image colour analysis; image enhancement; image retrieval; iterative methods; pattern clustering; rendering (computer graphics); visual databases; ISODATA clustering algorithm; color transfer; environment map image; illumination radiance; image color style; image database; interior virtual scenes; iterations; material redesign; representative viewpoint retrieval; scene rendering; Clustering algorithms; Image color analysis; Lighting; Materials; Photonics; Rendering (computer graphics); Standards; ISODATA clustering algorithms; interior scene; redesign material;
Conference_Titel :
Digital Home (ICDH), 2014 5th International Conference on
Conference_Location :
Guangzhou
Print_ISBN :
978-1-4799-4285-5
DOI :
10.1109/ICDH.2014.31