Convolutional Neural Networks for Traffic Sign Classification Using Enhanced Colours

V. E. Sathishkumar; C. Sharmila; S. Santhiya; M. Poongundran; S. Sanjeeth; S. Pranesh

doi:10.1007/978-3-031-27622-4_4

Convolutional Neural Networks for Traffic Sign Classification Using Enhanced Colours

V. E. Sathishkumar
, C. Sharmila
, S. Santhiya
, M. Poongundran
, S. Sanjeeth
, S. Pranesh

Hanyang University
Kongu Engineering College

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

Abstract

These days, traffic signs etched on the roadways improve traffic safety by notifying drivers of speed restrictions as well as other potential hazards such as deep curvy routes, upcoming road repairs, or pedestrian crossings. This study focuses on a three-stage real-time Traffic Sign Recognition and Classification system that includes image segmentation, traffic sign detection, and classification based on the input image. To extract red areas from an image, a colour enhancement approach is utilized. Convolutional Neural Networks (CNN) are used to determine the content of the traffic signals discovered during detection, classification, and identification. Results show that CNN model provides the highest accuracy of 99% in validation phase.

Original language	English
Title of host publication	Deep Sciences for Computing and Communications - First International Conference, IconDeepCom 2022, Revised Selected Papers
Editors	Kottilingam Kottursamy, Annie Uthra, Ali Kashif Bashir, Utku Kose
Publisher	Springer Science and Business Media Deutschland GmbH
Pages	34-43
Number of pages	10
ISBN (Print)	9783031276217
DOIs	https://doi.org/10.1007/978-3-031-27622-4_4
Publication status	Published - 2023
Externally published	Yes
Event	1st International Conference on Deep Sciences for Computing and Communications, IconDeepCom 2022 - Chennai, India Duration: 17 Mar 2022 → 18 Mar 2022

Publication series

Name	Communications in Computer and Information Science
Volume	1719 CCIS
ISSN (Print)	1865-0929
ISSN (Electronic)	1865-0937

Conference

Conference	1st International Conference on Deep Sciences for Computing and Communications, IconDeepCom 2022
Country/Territory	India
City	Chennai
Period	17/03/22 → 18/03/22

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 3 Good Health and Well-being
SDG 11 Sustainable Cities and Communities

Keywords

Convolutional Neural Network
Deep learning
Image classification
Traffic sign

Access to Document

10.1007/978-3-031-27622-4_4

Cite this

Sathishkumar, V. E., Sharmila, C., Santhiya, S., Poongundran, M., Sanjeeth, S., & Pranesh, S. (2023). Convolutional Neural Networks for Traffic Sign Classification Using Enhanced Colours. In K. Kottursamy, A. Uthra, A. K. Bashir, & U. Kose (Eds.), Deep Sciences for Computing and Communications - First International Conference, IconDeepCom 2022, Revised Selected Papers (pp. 34-43). (Communications in Computer and Information Science; Vol. 1719 CCIS). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-031-27622-4_4

Sathishkumar, V. E. ; Sharmila, C. ; Santhiya, S. et al. / Convolutional Neural Networks for Traffic Sign Classification Using Enhanced Colours. Deep Sciences for Computing and Communications - First International Conference, IconDeepCom 2022, Revised Selected Papers. editor / Kottilingam Kottursamy ; Annie Uthra ; Ali Kashif Bashir ; Utku Kose. Springer Science and Business Media Deutschland GmbH, 2023. pp. 34-43 (Communications in Computer and Information Science).

@inproceedings{2c3d7ee98ccf4c1aa62f17bb07129e21,

title = "Convolutional Neural Networks for Traffic Sign Classification Using Enhanced Colours",

abstract = "These days, traffic signs etched on the roadways improve traffic safety by notifying drivers of speed restrictions as well as other potential hazards such as deep curvy routes, upcoming road repairs, or pedestrian crossings. This study focuses on a three-stage real-time Traffic Sign Recognition and Classification system that includes image segmentation, traffic sign detection, and classification based on the input image. To extract red areas from an image, a colour enhancement approach is utilized. Convolutional Neural Networks (CNN) are used to determine the content of the traffic signals discovered during detection, classification, and identification. Results show that CNN model provides the highest accuracy of 99\% in validation phase.",

keywords = "Convolutional Neural Network, Deep learning, Image classification, Traffic sign",

author = "Sathishkumar, \{V. E.\} and C. Sharmila and S. Santhiya and M. Poongundran and S. Sanjeeth and S. Pranesh",

note = "Publisher Copyright: {\textcopyright} 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.; 1st International Conference on Deep Sciences for Computing and Communications, IconDeepCom 2022 ; Conference date: 17-03-2022 Through 18-03-2022",

year = "2023",

doi = "10.1007/978-3-031-27622-4\_4",

language = "English",

isbn = "9783031276217",

series = "Communications in Computer and Information Science",

publisher = "Springer Science and Business Media Deutschland GmbH",

pages = "34--43",

editor = "Kottilingam Kottursamy and Annie Uthra and Bashir, \{Ali Kashif\} and Utku Kose",

booktitle = "Deep Sciences for Computing and Communications - First International Conference, IconDeepCom 2022, Revised Selected Papers",

}

Sathishkumar, VE, Sharmila, C, Santhiya, S, Poongundran, M, Sanjeeth, S & Pranesh, S 2023, Convolutional Neural Networks for Traffic Sign Classification Using Enhanced Colours. in K Kottursamy, A Uthra, AK Bashir & U Kose (eds), Deep Sciences for Computing and Communications - First International Conference, IconDeepCom 2022, Revised Selected Papers. Communications in Computer and Information Science, vol. 1719 CCIS, Springer Science and Business Media Deutschland GmbH, pp. 34-43, 1st International Conference on Deep Sciences for Computing and Communications, IconDeepCom 2022, Chennai, India, 17/03/22. https://doi.org/10.1007/978-3-031-27622-4_4

Convolutional Neural Networks for Traffic Sign Classification Using Enhanced Colours. / Sathishkumar, V. E.; Sharmila, C.; Santhiya, S. et al.
Deep Sciences for Computing and Communications - First International Conference, IconDeepCom 2022, Revised Selected Papers. ed. / Kottilingam Kottursamy; Annie Uthra; Ali Kashif Bashir; Utku Kose. Springer Science and Business Media Deutschland GmbH, 2023. p. 34-43 (Communications in Computer and Information Science; Vol. 1719 CCIS).

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

TY - GEN

T1 - Convolutional Neural Networks for Traffic Sign Classification Using Enhanced Colours

AU - Sathishkumar, V. E.

AU - Sharmila, C.

AU - Santhiya, S.

AU - Poongundran, M.

AU - Sanjeeth, S.

AU - Pranesh, S.

PY - 2023

Y1 - 2023

N2 - These days, traffic signs etched on the roadways improve traffic safety by notifying drivers of speed restrictions as well as other potential hazards such as deep curvy routes, upcoming road repairs, or pedestrian crossings. This study focuses on a three-stage real-time Traffic Sign Recognition and Classification system that includes image segmentation, traffic sign detection, and classification based on the input image. To extract red areas from an image, a colour enhancement approach is utilized. Convolutional Neural Networks (CNN) are used to determine the content of the traffic signals discovered during detection, classification, and identification. Results show that CNN model provides the highest accuracy of 99% in validation phase.

AB - These days, traffic signs etched on the roadways improve traffic safety by notifying drivers of speed restrictions as well as other potential hazards such as deep curvy routes, upcoming road repairs, or pedestrian crossings. This study focuses on a three-stage real-time Traffic Sign Recognition and Classification system that includes image segmentation, traffic sign detection, and classification based on the input image. To extract red areas from an image, a colour enhancement approach is utilized. Convolutional Neural Networks (CNN) are used to determine the content of the traffic signals discovered during detection, classification, and identification. Results show that CNN model provides the highest accuracy of 99% in validation phase.

KW - Convolutional Neural Network

KW - Deep learning

KW - Image classification

KW - Traffic sign

UR - https://www.scopus.com/pages/publications/85151065738

U2 - 10.1007/978-3-031-27622-4_4

DO - 10.1007/978-3-031-27622-4_4

M3 - Conference contribution

AN - SCOPUS:85151065738

SN - 9783031276217

T3 - Communications in Computer and Information Science

SP - 34

EP - 43

BT - Deep Sciences for Computing and Communications - First International Conference, IconDeepCom 2022, Revised Selected Papers

A2 - Kottursamy, Kottilingam

A2 - Uthra, Annie

A2 - Bashir, Ali Kashif

A2 - Kose, Utku

PB - Springer Science and Business Media Deutschland GmbH

T2 - 1st International Conference on Deep Sciences for Computing and Communications, IconDeepCom 2022

Y2 - 17 March 2022 through 18 March 2022

ER -

Sathishkumar VE, Sharmila C, Santhiya S, Poongundran M, Sanjeeth S, Pranesh S. Convolutional Neural Networks for Traffic Sign Classification Using Enhanced Colours. In Kottursamy K, Uthra A, Bashir AK, Kose U, editors, Deep Sciences for Computing and Communications - First International Conference, IconDeepCom 2022, Revised Selected Papers. Springer Science and Business Media Deutschland GmbH. 2023. p. 34-43. (Communications in Computer and Information Science). doi: 10.1007/978-3-031-27622-4_4

Convolutional Neural Networks for Traffic Sign Classification Using Enhanced Colours

Abstract

Publication series

Conference

UN SDGs

Keywords

Access to Document

Other files and links

Fingerprint

Cite this