Real-time stereo matching on CUDA using an iterative refinement method for adaptive support-weight correspondences

Jeȩdrzej Kowalczuk, Eric T. Psota, Lance C Perez

Research output: Contribution to journalArticle

86 Citations (Scopus)

Abstract

High-quality real-time stereo matching has the potential to enable various computer vision applications including semi-automated robotic surgery, teleimmersion, and 3-D video surveillance. A novel real-time stereo matching method is presented that uses a two-pass approximation of adaptive support-weight aggregation, and a low-complexity iterative disparity refinement technique. Through an evaluation of computationally efficient approaches to adaptive support-weight cost aggregation, it is shown that the two-pass method produces an accurate approximation of the support weights while greatly reducing the complexity of aggregation. The refinement technique, constructed using a probabilistic framework, incorporates an additive term into matching cost minimization and facilitates iterative processing to improve the accuracy of the disparity map. This method has been implemented on massively parallel high-performance graphics hardware using the Compute Unified Device Architecture computing engine. Results show that the proposed method is the most accurate among all of the real-time stereo matching methods listed on the Middlebury stereo benchmark.

Original languageEnglish (US)
Article number6213098
Pages (from-to)94-104
Number of pages11
JournalIEEE Transactions on Circuits and Systems for Video Technology
Volume23
Issue number1
DOIs
StatePublished - Jan 21 2013

Fingerprint

Agglomeration
Computer vision
Costs
Engines
Hardware
Processing
Robotic surgery

Keywords

  • Adaptive support weights
  • CUDA
  • iterative refinement
  • real-time stereo matching

ASJC Scopus subject areas

  • Media Technology
  • Electrical and Electronic Engineering

Cite this

Real-time stereo matching on CUDA using an iterative refinement method for adaptive support-weight correspondences. / Kowalczuk, Jeȩdrzej; Psota, Eric T.; Perez, Lance C.

In: IEEE Transactions on Circuits and Systems for Video Technology, Vol. 23, No. 1, 6213098, 21.01.2013, p. 94-104.

Research output: Contribution to journalArticle

@article{641b4d827fd5443884c6d1735c2b040c,
title = "Real-time stereo matching on CUDA using an iterative refinement method for adaptive support-weight correspondences",
abstract = "High-quality real-time stereo matching has the potential to enable various computer vision applications including semi-automated robotic surgery, teleimmersion, and 3-D video surveillance. A novel real-time stereo matching method is presented that uses a two-pass approximation of adaptive support-weight aggregation, and a low-complexity iterative disparity refinement technique. Through an evaluation of computationally efficient approaches to adaptive support-weight cost aggregation, it is shown that the two-pass method produces an accurate approximation of the support weights while greatly reducing the complexity of aggregation. The refinement technique, constructed using a probabilistic framework, incorporates an additive term into matching cost minimization and facilitates iterative processing to improve the accuracy of the disparity map. This method has been implemented on massively parallel high-performance graphics hardware using the Compute Unified Device Architecture computing engine. Results show that the proposed method is the most accurate among all of the real-time stereo matching methods listed on the Middlebury stereo benchmark.",
keywords = "Adaptive support weights, CUDA, iterative refinement, real-time stereo matching",
author = "Jeȩdrzej Kowalczuk and Psota, {Eric T.} and Perez, {Lance C}",
year = "2013",
month = "1",
day = "21",
doi = "10.1109/TCSVT.2012.2203200",
language = "English (US)",
volume = "23",
pages = "94--104",
journal = "IEEE Transactions on Circuits and Systems for Video Technology",
issn = "1051-8215",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "1",

}

TY - JOUR

T1 - Real-time stereo matching on CUDA using an iterative refinement method for adaptive support-weight correspondences

AU - Kowalczuk, Jeȩdrzej

AU - Psota, Eric T.

AU - Perez, Lance C

PY - 2013/1/21

Y1 - 2013/1/21

N2 - High-quality real-time stereo matching has the potential to enable various computer vision applications including semi-automated robotic surgery, teleimmersion, and 3-D video surveillance. A novel real-time stereo matching method is presented that uses a two-pass approximation of adaptive support-weight aggregation, and a low-complexity iterative disparity refinement technique. Through an evaluation of computationally efficient approaches to adaptive support-weight cost aggregation, it is shown that the two-pass method produces an accurate approximation of the support weights while greatly reducing the complexity of aggregation. The refinement technique, constructed using a probabilistic framework, incorporates an additive term into matching cost minimization and facilitates iterative processing to improve the accuracy of the disparity map. This method has been implemented on massively parallel high-performance graphics hardware using the Compute Unified Device Architecture computing engine. Results show that the proposed method is the most accurate among all of the real-time stereo matching methods listed on the Middlebury stereo benchmark.

AB - High-quality real-time stereo matching has the potential to enable various computer vision applications including semi-automated robotic surgery, teleimmersion, and 3-D video surveillance. A novel real-time stereo matching method is presented that uses a two-pass approximation of adaptive support-weight aggregation, and a low-complexity iterative disparity refinement technique. Through an evaluation of computationally efficient approaches to adaptive support-weight cost aggregation, it is shown that the two-pass method produces an accurate approximation of the support weights while greatly reducing the complexity of aggregation. The refinement technique, constructed using a probabilistic framework, incorporates an additive term into matching cost minimization and facilitates iterative processing to improve the accuracy of the disparity map. This method has been implemented on massively parallel high-performance graphics hardware using the Compute Unified Device Architecture computing engine. Results show that the proposed method is the most accurate among all of the real-time stereo matching methods listed on the Middlebury stereo benchmark.

KW - Adaptive support weights

KW - CUDA

KW - iterative refinement

KW - real-time stereo matching

UR - http://www.scopus.com/inward/record.url?scp=84872319772&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84872319772&partnerID=8YFLogxK

U2 - 10.1109/TCSVT.2012.2203200

DO - 10.1109/TCSVT.2012.2203200

M3 - Article

AN - SCOPUS:84872319772

VL - 23

SP - 94

EP - 104

JO - IEEE Transactions on Circuits and Systems for Video Technology

JF - IEEE Transactions on Circuits and Systems for Video Technology

SN - 1051-8215

IS - 1

M1 - 6213098

ER -