GPU-ASIFT: A fast fully affine-invariant feature extraction algorithm

Valeriu Codreanu; Feng Dong; Baoquan Liu; Jos B.T.M. Roerdink; David Williams; Po Yang; Burhan Yasar

doi:10.1109/HPCSim.2013.6641456

GPU-ASIFT: A fast fully affine-invariant feature extraction algorithm

Valeriu Codreanu
, Feng Dong
, Baoquan Liu
, Jos B.T.M. Roerdink
, David Williams
, Po Yang
, Burhan Yasar

Institute for Research in Applicable Computing (IRAC)

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

26 Citations (Scopus)

Abstract

This paper presents a method that takes advantage of powerful graphics hardware to obtain fully affine-invariant image feature detection and matching. The chosen approach is the accurate, but also very computationally expensive, ASIFT algorithm. We have created a CUDA version of this algorithm that is up to 70 times faster than the original implementation, while keeping the algorithm's accuracy close to that of ASIFT. It's matching performance is therefore much better than that of other non-fully affine-invariant algorithms. Also, this approach was adapted to fit the multi-GPU paradigm in order to assess the acceleration potential from modern GPU clusters.

Original language	English
Title of host publication	Proceedings of the 2013 International Conference on High Performance Computing and Simulation, HPCS 2013
Pages	474-481
Number of pages	8
DOIs	https://doi.org/10.1109/HPCSim.2013.6641456
Publication status	Published - 21 Oct 2013
Event	2013 11th International Conference on High Performance Computing and Simulation, HPCS 2013 - Helsinki, Finland Duration: 1 Jul 2013 → 5 Jul 2013

Publication series

Name	Proceedings of the 2013 International Conference on High Performance Computing and Simulation, HPCS 2013

Conference

Conference	2013 11th International Conference on High Performance Computing and Simulation, HPCS 2013
Country/Territory	Finland
City	Helsinki
Period	1/07/13 → 5/07/13

Keywords

Computer Vision
GPU
High performance computing

ASJC Scopus subject areas

Applied Mathematics
Modeling and Simulation

Access to Document

10.1109/HPCSim.2013.6641456

https://ieeexplore.ieee.org/document/6641456

Handle.net

Persistent link

Cite this

Codreanu, V., Dong, F., Liu, B., Roerdink, J. B. T. M., Williams, D., Yang, P., & Yasar, B. (2013). GPU-ASIFT: A fast fully affine-invariant feature extraction algorithm. In Proceedings of the 2013 International Conference on High Performance Computing and Simulation, HPCS 2013 (pp. 474-481). Article 6641456 (Proceedings of the 2013 International Conference on High Performance Computing and Simulation, HPCS 2013). https://doi.org/10.1109/HPCSim.2013.6641456

@inproceedings{3475fbc28a094e15b14514b96e7b6caa,

title = "GPU-ASIFT: A fast fully affine-invariant feature extraction algorithm",

abstract = "This paper presents a method that takes advantage of powerful graphics hardware to obtain fully affine-invariant image feature detection and matching. The chosen approach is the accurate, but also very computationally expensive, ASIFT algorithm. We have created a CUDA version of this algorithm that is up to 70 times faster than the original implementation, while keeping the algorithm's accuracy close to that of ASIFT. It's matching performance is therefore much better than that of other non-fully affine-invariant algorithms. Also, this approach was adapted to fit the multi-GPU paradigm in order to assess the acceleration potential from modern GPU clusters.",

keywords = "Computer Vision, GPU, High performance computing",

author = "Valeriu Codreanu and Feng Dong and Baoquan Liu and Roerdink, \{Jos B.T.M.\} and David Williams and Po Yang and Burhan Yasar",

year = "2013",

month = oct,

day = "21",

doi = "10.1109/HPCSim.2013.6641456",

language = "English",

isbn = "9781479908363",

series = "Proceedings of the 2013 International Conference on High Performance Computing and Simulation, HPCS 2013",

pages = "474--481",

booktitle = "Proceedings of the 2013 International Conference on High Performance Computing and Simulation, HPCS 2013",

note = "2013 11th International Conference on High Performance Computing and Simulation, HPCS 2013 ; Conference date: 01-07-2013 Through 05-07-2013",

}

Codreanu, V, Dong, F, Liu, B, Roerdink, JBTM, Williams, D, Yang, P & Yasar, B 2013, GPU-ASIFT: A fast fully affine-invariant feature extraction algorithm. in Proceedings of the 2013 International Conference on High Performance Computing and Simulation, HPCS 2013., 6641456, Proceedings of the 2013 International Conference on High Performance Computing and Simulation, HPCS 2013, pp. 474-481, 2013 11th International Conference on High Performance Computing and Simulation, HPCS 2013, Helsinki, Finland, 1/07/13. https://doi.org/10.1109/HPCSim.2013.6641456

GPU-ASIFT: A fast fully affine-invariant feature extraction algorithm. / Codreanu, Valeriu; Dong, Feng; Liu, Baoquan et al.
Proceedings of the 2013 International Conference on High Performance Computing and Simulation, HPCS 2013. 2013. p. 474-481 6641456 (Proceedings of the 2013 International Conference on High Performance Computing and Simulation, HPCS 2013).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - GPU-ASIFT

T2 - 2013 11th International Conference on High Performance Computing and Simulation, HPCS 2013

AU - Codreanu, Valeriu

AU - Dong, Feng

AU - Liu, Baoquan

AU - Roerdink, Jos B.T.M.

AU - Williams, David

AU - Yang, Po

AU - Yasar, Burhan

PY - 2013/10/21

Y1 - 2013/10/21

N2 - This paper presents a method that takes advantage of powerful graphics hardware to obtain fully affine-invariant image feature detection and matching. The chosen approach is the accurate, but also very computationally expensive, ASIFT algorithm. We have created a CUDA version of this algorithm that is up to 70 times faster than the original implementation, while keeping the algorithm's accuracy close to that of ASIFT. It's matching performance is therefore much better than that of other non-fully affine-invariant algorithms. Also, this approach was adapted to fit the multi-GPU paradigm in order to assess the acceleration potential from modern GPU clusters.

AB - This paper presents a method that takes advantage of powerful graphics hardware to obtain fully affine-invariant image feature detection and matching. The chosen approach is the accurate, but also very computationally expensive, ASIFT algorithm. We have created a CUDA version of this algorithm that is up to 70 times faster than the original implementation, while keeping the algorithm's accuracy close to that of ASIFT. It's matching performance is therefore much better than that of other non-fully affine-invariant algorithms. Also, this approach was adapted to fit the multi-GPU paradigm in order to assess the acceleration potential from modern GPU clusters.

KW - Computer Vision

KW - GPU

KW - High performance computing

UR - https://www.scopus.com/pages/publications/84888022070

U2 - 10.1109/HPCSim.2013.6641456

DO - 10.1109/HPCSim.2013.6641456

M3 - Conference contribution

AN - SCOPUS:84888022070

SN - 9781479908363

T3 - Proceedings of the 2013 International Conference on High Performance Computing and Simulation, HPCS 2013

SP - 474

EP - 481

BT - Proceedings of the 2013 International Conference on High Performance Computing and Simulation, HPCS 2013

Y2 - 1 July 2013 through 5 July 2013

ER -

Codreanu V, Dong F, Liu B, Roerdink JBTM, Williams D, Yang P et al. GPU-ASIFT: A fast fully affine-invariant feature extraction algorithm. In Proceedings of the 2013 International Conference on High Performance Computing and Simulation, HPCS 2013. 2013. p. 474-481. 6641456. (Proceedings of the 2013 International Conference on High Performance Computing and Simulation, HPCS 2013). doi: 10.1109/HPCSim.2013.6641456

GPU-ASIFT: A fast fully affine-invariant feature extraction algorithm

Abstract

Publication series

Conference

Keywords

ASJC Scopus subject areas

Access to Document

Handle.net

Other files and links

Fingerprint

Cite this