Application of a Convolutional Neural Network for Detection of Ignition Sources and Smoke

Ilya R. Aliev; Vitalii A. Pavlov; Sergey V. Zavjalov; Yekaterina Sadovaya

doi:10.1007/978-3-030-58868-7_47

Application of a Convolutional Neural Network for Detection of Ignition Sources and Smoke

Ilya R. Aliev
, Vitalii A. Pavlov^*
, Sergey V. Zavjalov
, Yekaterina Sadovaya

^*Corresponding author for this work

Electrical Engineering

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

1 Citation (Scopus)

Abstract

The article discusses various methods for detecting ignition sources on aerial photographs. An algorithm based on color filtering and biorthogonal wavelet transform and the Tiny-YOLOv3 convolutional neural network were chosen for research. For the study, training and test datasets were developed. According to the results of experiments, Tiny-YOLOv3 exceeded the algorithm based on color filtering and biorthogonal wavelet transform in detection accuracy. For image processing algorithm AP was 16%. For the Tiny-YOLOv3 with input layer size of 416 × 416, the detection accuracy (AP) of fire and smoke was 56.5% and 31.9%, respectively.

Original language	English
Title of host publication	International Youth Conference on Electronics, Telecommunications and Information Technologies - Proceedings of the YETI 2020
Editors	Elena Velichko, Maksim Vinnichenko, Victoria Kapralova, Yevgeni Koucheryavy
Publisher	Springer
Pages	421-427
Number of pages	7
ISBN (Print)	9783030588670
DOIs	https://doi.org/10.1007/978-3-030-58868-7_47
Publication status	Published - 2021
Publication type	A4 Article in conference proceedings
Event	International Youth Conference on Electronics, Telecommunications and Information Technologies - St. Petersburg, Russian Federation Duration: 10 Jul 2020 → 11 Jul 2020

Publication series

Name	Springer Proceedings in Physics
Volume	255
ISSN (Print)	0930-8989
ISSN (Electronic)	1867-4941

Conference

Conference	International Youth Conference on Electronics, Telecommunications and Information Technologies
Abbreviated title	YETI
Country/Territory	Russian Federation
City	St. Petersburg
Period	10/07/20 → 11/07/20

Funding

Acknowledgments This research was funded by Peter the Great St. Petersburg Polytechnic University in the framework of the Program “5-100-2020” and used computational resources of Peter the Great Saint-Petersburg Polytechnic University Supercomputing Center (http://www.scc. spbstu.ru).

Keywords

Computer vision
Convolutional neural networks
Detection
Tiny-YOLOv3

Publication forum classification

Publication forum level 0

ASJC Scopus subject areas

General Physics and Astronomy

Access to Document

10.1007/978-3-030-58868-7_47

Cite this

Aliev, I. R., Pavlov, V. A., Zavjalov, S. V., & Sadovaya, Y. (2021). Application of a Convolutional Neural Network for Detection of Ignition Sources and Smoke. In E. Velichko, M. Vinnichenko, V. Kapralova, & Y. Koucheryavy (Eds.), International Youth Conference on Electronics, Telecommunications and Information Technologies - Proceedings of the YETI 2020 (pp. 421-427). (Springer Proceedings in Physics; Vol. 255). Springer. https://doi.org/10.1007/978-3-030-58868-7_47

Aliev, Ilya R. ; Pavlov, Vitalii A. ; Zavjalov, Sergey V. et al. / Application of a Convolutional Neural Network for Detection of Ignition Sources and Smoke. International Youth Conference on Electronics, Telecommunications and Information Technologies - Proceedings of the YETI 2020. editor / Elena Velichko ; Maksim Vinnichenko ; Victoria Kapralova ; Yevgeni Koucheryavy. Springer, 2021. pp. 421-427 (Springer Proceedings in Physics).

@inproceedings{afdb1eb50bfa4c3a8640f3d4b5f00068,

title = "Application of a Convolutional Neural Network for Detection of Ignition Sources and Smoke",

abstract = "The article discusses various methods for detecting ignition sources on aerial photographs. An algorithm based on color filtering and biorthogonal wavelet transform and the Tiny-YOLOv3 convolutional neural network were chosen for research. For the study, training and test datasets were developed. According to the results of experiments, Tiny-YOLOv3 exceeded the algorithm based on color filtering and biorthogonal wavelet transform in detection accuracy. For image processing algorithm AP was 16\%. For the Tiny-YOLOv3 with input layer size of 416 × 416, the detection accuracy (AP) of fire and smoke was 56.5\% and 31.9\%, respectively.",

keywords = "Computer vision, Convolutional neural networks, Detection, Tiny-YOLOv3",

author = "Aliev, \{Ilya R.\} and Pavlov, \{Vitalii A.\} and Zavjalov, \{Sergey V.\} and Yekaterina Sadovaya",

note = "jufoid=84327 Funding Information: Acknowledgments This research was funded by Peter the Great St. Petersburg Polytechnic University in the framework of the Program “5-100-2020” and used computational resources of Peter the Great Saint-Petersburg Polytechnic University Supercomputing Center (http://www.scc. spbstu.ru). Publisher Copyright: {\textcopyright} 2021, Springer Nature Switzerland AG. Copyright: Copyright 2020 Elsevier B.V., All rights reserved.; International Youth Conference on Electronics, Telecommunications and Information Technologies, YETI ; Conference date: 10-07-2020 Through 11-07-2020",

year = "2021",

doi = "10.1007/978-3-030-58868-7\_47",

language = "English",

isbn = "9783030588670",

series = "Springer Proceedings in Physics",

publisher = "Springer",

pages = "421--427",

editor = "Elena Velichko and Maksim Vinnichenko and Victoria Kapralova and Yevgeni Koucheryavy",

booktitle = "International Youth Conference on Electronics, Telecommunications and Information Technologies - Proceedings of the YETI 2020",

}

Aliev, IR, Pavlov, VA, Zavjalov, SV & Sadovaya, Y 2021, Application of a Convolutional Neural Network for Detection of Ignition Sources and Smoke. in E Velichko, M Vinnichenko, V Kapralova & Y Koucheryavy (eds), International Youth Conference on Electronics, Telecommunications and Information Technologies - Proceedings of the YETI 2020. Springer Proceedings in Physics, vol. 255, Springer, pp. 421-427, International Youth Conference on Electronics, Telecommunications and Information Technologies, St. Petersburg, Russian Federation, 10/07/20. https://doi.org/10.1007/978-3-030-58868-7_47

Application of a Convolutional Neural Network for Detection of Ignition Sources and Smoke. / Aliev, Ilya R.; Pavlov, Vitalii A.; Zavjalov, Sergey V. et al.
International Youth Conference on Electronics, Telecommunications and Information Technologies - Proceedings of the YETI 2020. ed. / Elena Velichko; Maksim Vinnichenko; Victoria Kapralova; Yevgeni Koucheryavy. Springer, 2021. p. 421-427 (Springer Proceedings in Physics; Vol. 255).

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

TY - GEN

T1 - Application of a Convolutional Neural Network for Detection of Ignition Sources and Smoke

AU - Aliev, Ilya R.

AU - Pavlov, Vitalii A.

AU - Zavjalov, Sergey V.

AU - Sadovaya, Yekaterina

N1 - jufoid=84327 Funding Information: Acknowledgments This research was funded by Peter the Great St. Petersburg Polytechnic University in the framework of the Program “5-100-2020” and used computational resources of Peter the Great Saint-Petersburg Polytechnic University Supercomputing Center (http://www.scc. spbstu.ru). Publisher Copyright: © 2021, Springer Nature Switzerland AG. Copyright: Copyright 2020 Elsevier B.V., All rights reserved.

PY - 2021

Y1 - 2021

N2 - The article discusses various methods for detecting ignition sources on aerial photographs. An algorithm based on color filtering and biorthogonal wavelet transform and the Tiny-YOLOv3 convolutional neural network were chosen for research. For the study, training and test datasets were developed. According to the results of experiments, Tiny-YOLOv3 exceeded the algorithm based on color filtering and biorthogonal wavelet transform in detection accuracy. For image processing algorithm AP was 16%. For the Tiny-YOLOv3 with input layer size of 416 × 416, the detection accuracy (AP) of fire and smoke was 56.5% and 31.9%, respectively.

AB - The article discusses various methods for detecting ignition sources on aerial photographs. An algorithm based on color filtering and biorthogonal wavelet transform and the Tiny-YOLOv3 convolutional neural network were chosen for research. For the study, training and test datasets were developed. According to the results of experiments, Tiny-YOLOv3 exceeded the algorithm based on color filtering and biorthogonal wavelet transform in detection accuracy. For image processing algorithm AP was 16%. For the Tiny-YOLOv3 with input layer size of 416 × 416, the detection accuracy (AP) of fire and smoke was 56.5% and 31.9%, respectively.

KW - Computer vision

KW - Convolutional neural networks

KW - Detection

KW - Tiny-YOLOv3

U2 - 10.1007/978-3-030-58868-7_47

DO - 10.1007/978-3-030-58868-7_47

M3 - Conference contribution

AN - SCOPUS:85097652852

SN - 9783030588670

T3 - Springer Proceedings in Physics

SP - 421

EP - 427

BT - International Youth Conference on Electronics, Telecommunications and Information Technologies - Proceedings of the YETI 2020

A2 - Velichko, Elena

A2 - Vinnichenko, Maksim

A2 - Kapralova, Victoria

A2 - Koucheryavy, Yevgeni

PB - Springer

T2 - International Youth Conference on Electronics, Telecommunications and Information Technologies

Y2 - 10 July 2020 through 11 July 2020

ER -

Aliev IR, Pavlov VA, Zavjalov SV, Sadovaya Y. Application of a Convolutional Neural Network for Detection of Ignition Sources and Smoke. In Velichko E, Vinnichenko M, Kapralova V, Koucheryavy Y, editors, International Youth Conference on Electronics, Telecommunications and Information Technologies - Proceedings of the YETI 2020. Springer. 2021. p. 421-427. (Springer Proceedings in Physics). Epub 2020. doi: 10.1007/978-3-030-58868-7_47

Application of a Convolutional Neural Network for Detection of Ignition Sources and Smoke

Abstract

Publication series

Conference

Funding

Keywords

Publication forum classification

ASJC Scopus subject areas

Access to Document

Fingerprint

Cite this