DocumentCode :
3270192
Title :
Furry stylized texel-rendering in images and videos
Author :
Jiao, Shaohui ; Sheng, Bin ; Sun, Hanqiu ; Wu, Enhua
Author_Institution :
Inst. of Software, Chinese Acad. of Sci., Beijing, China
fYear :
2009
fDate :
8-10 Dec. 2009
Firstpage :
1
Lastpage :
5
Abstract :
We present a new method for adding furry effects for cartoon characters in images and videos. We synthesize furry stylized textures based on 3D texel structure. Given an image or a video as input, realistic fur texels are generated and mapped onto the region of interest(ROI) which is obtained using the Snake method. The 3D texels are rendered using ray marching algorithm on GPU for fast synthesizing the furry stylized textures. Specifically, fur texels of triangular prism shape are generated in order to be possibly mapped onto the traditional triangular mesh. Gaussian function is also introduced to provide parameters for users to control the fuzz degree. For the video applications, we use SIFT(scale invariant feature transform) algorithm to extract feature points and calculate the feature matching of the ROI along with the timeline. Experimental results show that the presented method is able to synthesize the furry stylized textures in the input image and generate dynamic fur in a video for better realistic rendering.
Keywords :
image texture; rendering (computer graphics); transforms; video signal processing; Furry stylized texel-rendering; Gaussian function; Snake method; cartoon characters; furry stylized textures; fuzz degree; ray marching algorithm; region of interest; scale invariant feature transform; Animals; Computer science; Feature extraction; Image generation; Image sequences; Mesh generation; Rendering (computer graphics); Shape control; Sun; Videos; 3D texels; SIFT; furry surface; image-based rendering;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Information, Communications and Signal Processing, 2009. ICICS 2009. 7th International Conference on
Conference_Location :
Macau
Print_ISBN :
978-1-4244-4656-8
Electronic_ISBN :
978-1-4244-4657-5
Type :
conf
DOI :
10.1109/ICICS.2009.5397588
Filename :
5397588
Link To Document :
بازگشت