TY - GEN
T1 - A highly parallel design of image surface layout recovering on GPGPU
AU - Li, Guan Ru
AU - Lai, Bo-Cheng
PY - 2012
Y1 - 2012
N2 - 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.
AB - 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.
UR - http://www.scopus.com/inward/record.url?scp=84864032897&partnerID=8YFLogxK
U2 - 10.1109/VLSI-DAT.2012.6212628
DO - 10.1109/VLSI-DAT.2012.6212628
M3 - Conference contribution
AN - SCOPUS:84864032897
SN - 9781457720819
T3 - 2012 International Symposium on VLSI Design, Automation and Test, VLSI-DAT 2012 - Proceedings of Technical Papers
BT - 2012 International Symposium on VLSI Design, Automation and Test, VLSI-DAT 2012 - Proceedings of Technical Papers
T2 - 2012 International Symposium on VLSI Design, Automation and Test, VLSI-DAT 2012
Y2 - 23 April 2012 through 25 April 2012
ER -