A highly parallel design of image surface layout recovering on GPGPU

Author

Li, Guan-Ru ; Lai, Bo-Cheng Charles

Author_Institution

Dept. of Electron. Eng., Nat. Chiao Tung Univ., Hsinchu, Taiwan

fYear

2012

fDate

23-25 April 2012

Firstpage

1

Lastpage

4

Abstract

Surface layout recovering helps computers understand the intricate information in an image by assigning local segments to different geometric classes. It greatly reduces the complexity of the following-up image processing and is widely used in various computer vision applications. However, the algorithm walks through every image pixel and imposes intensive computation requirement. Through comprehensive analysis on the execution behavior, this paper identifies significant parallelism inherent in the algorithm. With careful concerns on both multi-threaded software and parallel hardware, the optimized parallel design on a modern GPGPU has reached an average of 10.7X performance enhancement.

Keywords

computer vision; graphics processing units; multi-threading; GPGPU; computer vision applications; general purpose graphic processing units; image information; image pixel; image processing; image surface layout recovery; multithreaded software; optimized parallel design; Algorithm design and analysis; Hardware; Image segmentation; Instruction sets; Layout; Parallel processing; Surface treatment;

fLanguage

English

Publisher

ieee

Conference_Titel

VLSI Design, Automation, and Test (VLSI-DAT), 2012 International Symposium on

Conference_Location

Hsinchu

ISSN

PENDING

Print_ISBN

978-1-4577-2080-2

Type

conf

DOI

10.1109/VLSI-DAT.2012.6212628

Filename

6212628