Parallel realization of SIFT feature abstraction based on GPU using Tegra K1

Donglin Jiang; Jianzhi Long; Wanwan Yu; Ye Jin

Parallel realization of SIFT feature abstraction based on GPU using Tegra K1

Donglin Jiang, Jianzhi Long, Wanwan Yu, Ye Jin

School of Information and Electronics

Research output: Contribution to conference › Paper › peer-review

Abstract

The fast and steady feature extraction and key point matching are key factors to SAR (synthetic aperture radar) image automatic registration algorithm. The real-time processing of the algorithm in SAR is one of the challenges that need to be solved urgently. This paper aims at improving the efficiency of the description of key point in SIFT algorithm. We implemented the rapid calculation of the feature extraction section in SIFT (scale invariant feature transform) based on embedded NVIDIA GPU-Tegra K1 by fully taking the advantages of multi-cores of GPU in parallel computing, floating point calculation, memory management, etc. And the speed ratio of the accelerated calculation of the SIFT feature reaches 5.5.

Original language	English
Publication status	Published - 2015
Event	IET International Radar Conference 2015 - Hangzhou, China Duration: 14 Oct 2015 → 16 Oct 2015

Conference

Conference	IET International Radar Conference 2015
Country/Territory	China
City	Hangzhou
Period	14/10/15 → 16/10/15

Keywords

NVIDIA Tegra K1
Parallel computing
SAR image registration
SIFT feature extraction

Cite this

@conference{a53d225102ad47adbc824052bf44aea2,

title = "Parallel realization of SIFT feature abstraction based on GPU using Tegra K1",

abstract = "The fast and steady feature extraction and key point matching are key factors to SAR (synthetic aperture radar) image automatic registration algorithm. The real-time processing of the algorithm in SAR is one of the challenges that need to be solved urgently. This paper aims at improving the efficiency of the description of key point in SIFT algorithm. We implemented the rapid calculation of the feature extraction section in SIFT (scale invariant feature transform) based on embedded NVIDIA GPU-Tegra K1 by fully taking the advantages of multi-cores of GPU in parallel computing, floating point calculation, memory management, etc. And the speed ratio of the accelerated calculation of the SIFT feature reaches 5.5.",

keywords = "NVIDIA Tegra K1, Parallel computing, SAR image registration, SIFT feature extraction",

author = "Donglin Jiang and Jianzhi Long and Wanwan Yu and Ye Jin",

year = "2015",

language = "English",

note = "IET International Radar Conference 2015 ; Conference date: 14-10-2015 Through 16-10-2015",

}

TY - CONF

T1 - Parallel realization of SIFT feature abstraction based on GPU using Tegra K1

AU - Jiang, Donglin

AU - Long, Jianzhi

AU - Yu, Wanwan

AU - Jin, Ye

PY - 2015

Y1 - 2015

N2 - The fast and steady feature extraction and key point matching are key factors to SAR (synthetic aperture radar) image automatic registration algorithm. The real-time processing of the algorithm in SAR is one of the challenges that need to be solved urgently. This paper aims at improving the efficiency of the description of key point in SIFT algorithm. We implemented the rapid calculation of the feature extraction section in SIFT (scale invariant feature transform) based on embedded NVIDIA GPU-Tegra K1 by fully taking the advantages of multi-cores of GPU in parallel computing, floating point calculation, memory management, etc. And the speed ratio of the accelerated calculation of the SIFT feature reaches 5.5.

AB - The fast and steady feature extraction and key point matching are key factors to SAR (synthetic aperture radar) image automatic registration algorithm. The real-time processing of the algorithm in SAR is one of the challenges that need to be solved urgently. This paper aims at improving the efficiency of the description of key point in SIFT algorithm. We implemented the rapid calculation of the feature extraction section in SIFT (scale invariant feature transform) based on embedded NVIDIA GPU-Tegra K1 by fully taking the advantages of multi-cores of GPU in parallel computing, floating point calculation, memory management, etc. And the speed ratio of the accelerated calculation of the SIFT feature reaches 5.5.

KW - NVIDIA Tegra K1

KW - Parallel computing

KW - SAR image registration

KW - SIFT feature extraction

UR - http://www.scopus.com/inward/record.url?scp=84973582561&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:84973582561

T2 - IET International Radar Conference 2015

Y2 - 14 October 2015 through 16 October 2015

ER -

Parallel realization of SIFT feature abstraction based on GPU using Tegra K1

Abstract

Conference

Keywords

Other files and links

Fingerprint

Cite this