Solving the stereo matching problem using an embedded GPU for a real-Time application

Pedro Aguiar, Sebastien Varrier, Jorge Lozoya, Martha Lopez, Damien Koenig, Juan C. Tudon-Martinez

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

General-purpose computing on graphics processing units (GPGPU) is used not only to offload the CPU from heavy computations but also to perform them faster than it is possible on CPUs. This is commonly referred as GPU acceleration and is an exercised area of study in the PC platform that has received very little attention on commodity embedded devices. Just as the PC GPU is being used to perform computations that would be impossible in terms of execution time for its accompanying CPU, the embedded GPU can accelerate computations normally done by the embedded CPU. This work presents an implementation of a factible, real-Time, GPU accelerated stereo matching solution using a Broadcom's VideoCore IV GPU (BCM2835 System on a Chip). Details include the delimitation process, design considerations and optimization techniques.

Original languageEnglish
Title of host publicationInternational Conference on Advanced Mechatronic Systems, ICAMechS
PublisherIEEE Computer Society
Pages519-524
Number of pages6
ISBN (Electronic)978-153862602-3
ISBN (Print)9781538626023
DOIs
Publication statusPublished - 14 Mar 2018
EventInternational Conference on Advanced Mechatronic Systems, ICAMechS -
Duration: 14 Mar 2018 → …

Publication series

NameInternational Conference on Advanced Mechatronic Systems, ICAMechS
Volume2017-December
ISSN (Print)2325-0682
ISSN (Electronic)2325-0690

Conference

ConferenceInternational Conference on Advanced Mechatronic Systems, ICAMechS
Period14/3/18 → …

Fingerprint

Program processors
Graphics processing unit
Process design

All Science Journal Classification (ASJC) codes

  • Mechanical Engineering
  • Electrical and Electronic Engineering

Cite this

Aguiar, P., Varrier, S., Lozoya, J., Lopez, M., Koenig, D., & Tudon-Martinez, J. C. (2018). Solving the stereo matching problem using an embedded GPU for a real-Time application. In International Conference on Advanced Mechatronic Systems, ICAMechS (pp. 519-524). (International Conference on Advanced Mechatronic Systems, ICAMechS; Vol. 2017-December). IEEE Computer Society. https://doi.org/10.1109/ICAMechS.2017.8316471
Aguiar, Pedro ; Varrier, Sebastien ; Lozoya, Jorge ; Lopez, Martha ; Koenig, Damien ; Tudon-Martinez, Juan C. / Solving the stereo matching problem using an embedded GPU for a real-Time application. International Conference on Advanced Mechatronic Systems, ICAMechS. IEEE Computer Society, 2018. pp. 519-524 (International Conference on Advanced Mechatronic Systems, ICAMechS).
@inproceedings{5f665624724544a7895fd57cf8bc6e07,
title = "Solving the stereo matching problem using an embedded GPU for a real-Time application",
abstract = "General-purpose computing on graphics processing units (GPGPU) is used not only to offload the CPU from heavy computations but also to perform them faster than it is possible on CPUs. This is commonly referred as GPU acceleration and is an exercised area of study in the PC platform that has received very little attention on commodity embedded devices. Just as the PC GPU is being used to perform computations that would be impossible in terms of execution time for its accompanying CPU, the embedded GPU can accelerate computations normally done by the embedded CPU. This work presents an implementation of a factible, real-Time, GPU accelerated stereo matching solution using a Broadcom's VideoCore IV GPU (BCM2835 System on a Chip). Details include the delimitation process, design considerations and optimization techniques.",
author = "Pedro Aguiar and Sebastien Varrier and Jorge Lozoya and Martha Lopez and Damien Koenig and Tudon-Martinez, {Juan C.}",
year = "2018",
month = "3",
day = "14",
doi = "10.1109/ICAMechS.2017.8316471",
language = "English",
isbn = "9781538626023",
series = "International Conference on Advanced Mechatronic Systems, ICAMechS",
publisher = "IEEE Computer Society",
pages = "519--524",
booktitle = "International Conference on Advanced Mechatronic Systems, ICAMechS",
address = "United States",

}

Aguiar, P, Varrier, S, Lozoya, J, Lopez, M, Koenig, D & Tudon-Martinez, JC 2018, Solving the stereo matching problem using an embedded GPU for a real-Time application. in International Conference on Advanced Mechatronic Systems, ICAMechS. International Conference on Advanced Mechatronic Systems, ICAMechS, vol. 2017-December, IEEE Computer Society, pp. 519-524, International Conference on Advanced Mechatronic Systems, ICAMechS, 14/3/18. https://doi.org/10.1109/ICAMechS.2017.8316471

Solving the stereo matching problem using an embedded GPU for a real-Time application. / Aguiar, Pedro; Varrier, Sebastien; Lozoya, Jorge; Lopez, Martha; Koenig, Damien; Tudon-Martinez, Juan C.

International Conference on Advanced Mechatronic Systems, ICAMechS. IEEE Computer Society, 2018. p. 519-524 (International Conference on Advanced Mechatronic Systems, ICAMechS; Vol. 2017-December).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Solving the stereo matching problem using an embedded GPU for a real-Time application

AU - Aguiar, Pedro

AU - Varrier, Sebastien

AU - Lozoya, Jorge

AU - Lopez, Martha

AU - Koenig, Damien

AU - Tudon-Martinez, Juan C.

PY - 2018/3/14

Y1 - 2018/3/14

N2 - General-purpose computing on graphics processing units (GPGPU) is used not only to offload the CPU from heavy computations but also to perform them faster than it is possible on CPUs. This is commonly referred as GPU acceleration and is an exercised area of study in the PC platform that has received very little attention on commodity embedded devices. Just as the PC GPU is being used to perform computations that would be impossible in terms of execution time for its accompanying CPU, the embedded GPU can accelerate computations normally done by the embedded CPU. This work presents an implementation of a factible, real-Time, GPU accelerated stereo matching solution using a Broadcom's VideoCore IV GPU (BCM2835 System on a Chip). Details include the delimitation process, design considerations and optimization techniques.

AB - General-purpose computing on graphics processing units (GPGPU) is used not only to offload the CPU from heavy computations but also to perform them faster than it is possible on CPUs. This is commonly referred as GPU acceleration and is an exercised area of study in the PC platform that has received very little attention on commodity embedded devices. Just as the PC GPU is being used to perform computations that would be impossible in terms of execution time for its accompanying CPU, the embedded GPU can accelerate computations normally done by the embedded CPU. This work presents an implementation of a factible, real-Time, GPU accelerated stereo matching solution using a Broadcom's VideoCore IV GPU (BCM2835 System on a Chip). Details include the delimitation process, design considerations and optimization techniques.

UR - http://www.scopus.com/inward/record.url?scp=85046679575&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85046679575&partnerID=8YFLogxK

U2 - 10.1109/ICAMechS.2017.8316471

DO - 10.1109/ICAMechS.2017.8316471

M3 - Conference contribution

SN - 9781538626023

T3 - International Conference on Advanced Mechatronic Systems, ICAMechS

SP - 519

EP - 524

BT - International Conference on Advanced Mechatronic Systems, ICAMechS

PB - IEEE Computer Society

ER -

Aguiar P, Varrier S, Lozoya J, Lopez M, Koenig D, Tudon-Martinez JC. Solving the stereo matching problem using an embedded GPU for a real-Time application. In International Conference on Advanced Mechatronic Systems, ICAMechS. IEEE Computer Society. 2018. p. 519-524. (International Conference on Advanced Mechatronic Systems, ICAMechS). https://doi.org/10.1109/ICAMechS.2017.8316471