DocumentCode
624602
Title
Implementation of face cartoon maker system based on android
Author
Kairong Wang
Author_Institution
Coll. of Comput. Sci. & Technol., Huazhong Univ. of Sci. & Technol., Wuhan, China
fYear
2013
fDate
9-11 June 2013
Firstpage
193
Lastpage
198
Abstract
In the field of non-photorealistic face generation for digital entertainment, face cartoon generation becomes a hot spot. Although all kinds of face cartoon software emerge in endlessly, most of them are based on PC platform and some need manual operation. In this paper, we present a Face Cartoon Maker System on android by improving the classical ASM (Active Shape Model) algorithm. Compared with the classical ASM, the improved ASM is extended by selectively using two-instead of one-dimensional landmark profiles. In system implementation, we first use Java JNI technology and OpenCV transplantation technology to transplant face detectors to from PC platform to android platform, then we use the improved ASM algorithm to extract facial features and at last generate the cartoon face automatically by Android components. Finally, on the base of ORL face database, we design a similarity analysis between cartoon face and realistic face among four different postures, and the result shows the recognition rate of the normal posture is the highest.
Keywords
Java; Linux; computer animation; face recognition; feature extraction; public domain software; realistic images; Android; Java JNI technology; ORL face database; OpenCV transplantation technology; digital entertainment; face cartoon generation; face cartoon maker system; face cartoon software; face detectors; facial feature extraction; improved ASM algorithm; improved active shape model algorithm; nonphotorealistic face generation; normal posture recognition rate; realistic face; similarity analysis; two-landmark profiles; Androids; Detectors; Face; Feature extraction; Java; Shape; Training;
fLanguage
English
Publisher
ieee
Conference_Titel
Intelligent Control and Information Processing (ICICIP), 2013 Fourth International Conference on
Conference_Location
Beijing
Print_ISBN
978-1-4673-6248-1
Type
conf
DOI
10.1109/ICICIP.2013.6568066
Filename
6568066
Link To Document