Real-time traffic sign recognition based on a general purpose GPU and deep-learning

Kwangyong Lim, Yongwon Hong, Yeongwoo Choi, Hyeran Byun

Research output: Contribution to journalArticle

12 Citations (Scopus)

Abstract

We present a General Purpose Graphics Processing Unit (GPGPU) based real-time traffic sign detection and recognition method that is robust against illumination changes. There have been many approaches to traffic sign recognition in various research fields; however, previous approaches faced several limitations when under low illumination or wide variance of light conditions. To overcome these drawbacks and improve processing speeds, we propose a method that 1) is robust against illumination changes, 2) uses GPGPU-based realtime traffic sign detection, and 3) performs region detecting and recognition using a hierarchical model. This method produces stable results in low illumination environments. Both detection and hierarchical recognition are performed in real-time, and the proposed method achieves 0.97 F1-score on our collective dataset, which uses the Vienna convention traffic rules (Germany and South Korea).

Original languageEnglish
Article numbere0173317
JournalPLoS One
Volume12
Issue number3
DOIs
Publication statusPublished - 2017 Mar 1

Fingerprint

Traffic signs
Lighting
traffic
lighting
learning
Learning
Republic of Korea
South Korea
methodology
Austria
Germany
Light
Recognition (Psychology)
Deep learning
Graphics processing unit
Processing
Research

All Science Journal Classification (ASJC) codes

  • Biochemistry, Genetics and Molecular Biology(all)
  • Agricultural and Biological Sciences(all)

Cite this

Lim, Kwangyong ; Hong, Yongwon ; Choi, Yeongwoo ; Byun, Hyeran. / Real-time traffic sign recognition based on a general purpose GPU and deep-learning. In: PLoS One. 2017 ; Vol. 12, No. 3.
@article{a234ded807234b508df7de9660e9f989,
title = "Real-time traffic sign recognition based on a general purpose GPU and deep-learning",
abstract = "We present a General Purpose Graphics Processing Unit (GPGPU) based real-time traffic sign detection and recognition method that is robust against illumination changes. There have been many approaches to traffic sign recognition in various research fields; however, previous approaches faced several limitations when under low illumination or wide variance of light conditions. To overcome these drawbacks and improve processing speeds, we propose a method that 1) is robust against illumination changes, 2) uses GPGPU-based realtime traffic sign detection, and 3) performs region detecting and recognition using a hierarchical model. This method produces stable results in low illumination environments. Both detection and hierarchical recognition are performed in real-time, and the proposed method achieves 0.97 F1-score on our collective dataset, which uses the Vienna convention traffic rules (Germany and South Korea).",
author = "Kwangyong Lim and Yongwon Hong and Yeongwoo Choi and Hyeran Byun",
year = "2017",
month = "3",
day = "1",
doi = "10.1371/journal.pone.0173317",
language = "English",
volume = "12",
journal = "PLoS One",
issn = "1932-6203",
publisher = "Public Library of Science",
number = "3",

}

Real-time traffic sign recognition based on a general purpose GPU and deep-learning. / Lim, Kwangyong; Hong, Yongwon; Choi, Yeongwoo; Byun, Hyeran.

In: PLoS One, Vol. 12, No. 3, e0173317, 01.03.2017.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Real-time traffic sign recognition based on a general purpose GPU and deep-learning

AU - Lim, Kwangyong

AU - Hong, Yongwon

AU - Choi, Yeongwoo

AU - Byun, Hyeran

PY - 2017/3/1

Y1 - 2017/3/1

N2 - We present a General Purpose Graphics Processing Unit (GPGPU) based real-time traffic sign detection and recognition method that is robust against illumination changes. There have been many approaches to traffic sign recognition in various research fields; however, previous approaches faced several limitations when under low illumination or wide variance of light conditions. To overcome these drawbacks and improve processing speeds, we propose a method that 1) is robust against illumination changes, 2) uses GPGPU-based realtime traffic sign detection, and 3) performs region detecting and recognition using a hierarchical model. This method produces stable results in low illumination environments. Both detection and hierarchical recognition are performed in real-time, and the proposed method achieves 0.97 F1-score on our collective dataset, which uses the Vienna convention traffic rules (Germany and South Korea).

AB - We present a General Purpose Graphics Processing Unit (GPGPU) based real-time traffic sign detection and recognition method that is robust against illumination changes. There have been many approaches to traffic sign recognition in various research fields; however, previous approaches faced several limitations when under low illumination or wide variance of light conditions. To overcome these drawbacks and improve processing speeds, we propose a method that 1) is robust against illumination changes, 2) uses GPGPU-based realtime traffic sign detection, and 3) performs region detecting and recognition using a hierarchical model. This method produces stable results in low illumination environments. Both detection and hierarchical recognition are performed in real-time, and the proposed method achieves 0.97 F1-score on our collective dataset, which uses the Vienna convention traffic rules (Germany and South Korea).

UR - http://www.scopus.com/inward/record.url?scp=85014734268&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85014734268&partnerID=8YFLogxK

U2 - 10.1371/journal.pone.0173317

DO - 10.1371/journal.pone.0173317

M3 - Article

VL - 12

JO - PLoS One

JF - PLoS One

SN - 1932-6203

IS - 3

M1 - e0173317

ER -