DocumentCode :
579333
Title :
Implementation and optimization of software depth estimation for arm
Author :
Tse Kai Heng ; Kawashima, Yoshiya ; Fujisawa, T. ; Yasuda, G. ; Oshikiri, M. ; Tanaka, T. ; Matsuoka, Koichi ; Kikuchi, Yutaka
Author_Institution :
Platform & Solution Dev. Center, Toshiba Corp., Tokyo, Japan
fYear :
2012
fDate :
15-17 Oct. 2012
Firstpage :
1
Lastpage :
4
Abstract :
3D imaging capable devices have been available on the market for a few years now and the availability of 3D contents is also picking up. However, 3D contents available for viewing is not growing fast enough to meet the demand. To compensate for such a problem, most 3D imaging devices are equipped with 2D to 3D conversion function. Also, common 3D contents available are mostly produced with 2 view points which poses a problem for autostereoscopic devices that is capable of displaying multiple view points. In the absence of multiview 3D contents, to produce natural viewing experience on such device, depth estimation is an essential technology in order to create more parallax images from source images. In this paper, we present the implementation of Toshiba´s software depth estimation process on ARM processor which allows realtime processing on embedded devices such as TVs, mobile phones and tablet computers adopting stereoscopic or autostereoscopic 3D technology.
Keywords :
embedded systems; image processing; optimisation; 3D contents; 3D imaging capable devices; ARM processor; embedded devices; optimization; parallax images; software depth estimation; Delay; Estimation; Multicore processing; Optimization; Pipelines; Software; Vectors; Consumer products; Embedded software; Image processing; Multicore processing; Optimization; Real time systems; System-on-a-chip; Three dimensional TV;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
3DTV-Conference: The True Vision - Capture, Transmission and Display of 3D Video (3DTV-CON), 2012
Conference_Location :
Zurich
ISSN :
2161-2021
Print_ISBN :
978-1-4673-4904-8
Electronic_ISBN :
2161-2021
Type :
conf
DOI :
10.1109/3DTV.2012.6365427
Filename :
6365427
Link To Document :
بازگشت