Research of neural network algorithms for recognizing railway infrastructure objects in video images

The article describes the development of two neural network algorithms for recognizing objects of the railway infrastructure in video images. Both algorithms are aimed at improving railway traffic safety. One algorithm detects foreign objects on railway tracks and objects relating to the railway inf...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:Kompʹûternaâ optika Ročník 49; číslo 3; s. 443 - 450
Hlavní autori: Medvedeva, E.V., Perevoshchikova, A.A.
Médium: Journal Article
Jazyk:English
Vydavateľské údaje: Samara National Research University 01.06.2025
Predmet:
machine vision systems
neural network algorithms
object detection
railway infrastructure objects
railway traffic safety
semantic segmentation
ISSN:0134-2452, 2412-6179
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Abstract The article describes the development of two neural network algorithms for recognizing objects of the railway infrastructure in video images. Both algorithms are aimed at improving railway traffic safety. One algorithm detects foreign objects on railway tracks and objects relating to the railway infrastructure. The other algorithm implements the semantic segmentation of main and auxiliary railway tracks, as well as trains within the visible range of the locomotive. The algorithms are implemented based on convolutional neural networks (CNN) YOLO and U-Net. The CNN is trained and tested using the image database of the Research Institute of Information, Automation and Communications in Railway Transport. The experimental studies conducted are aimed at increasing the efficiency of algorithms for object detection and segmentation through the use of data augmentation methods and additional preprocessing, as well as selecting an architecture and optimal network hyperparameters. The detection algorithm works in real time, achieving an average accuracy of 64% for 11 object classes according to the mAP metric. The operating speed of the semantic segmentation algorithm is 5 frames/s, the average accuracy for three classes of objects according to the IoU metric is 92%.
AbstractList The article describes the development of two neural network algorithms for recognizing objects of the railway infrastructure in video images. Both algorithms are aimed at improving railway traffic safety. One algorithm detects foreign objects on railway tracks and objects relating to the railway infrastructure. The other algorithm implements the semantic segmentation of main and auxiliary railway tracks, as well as trains within the visible range of the locomotive. The algorithms are implemented based on convolutional neural networks (CNN) YOLO and U-Net. The CNN is trained and tested using the image database of the Research Institute of Information, Automation and Communications in Railway Transport. The experimental studies conducted are aimed at increasing the efficiency of algorithms for object detection and segmentation through the use of data augmentation methods and additional preprocessing, as well as selecting an architecture and optimal network hyperparameters. The detection algorithm works in real time, achieving an average accuracy of 64% for 11 object classes according to the mAP metric. The operating speed of the semantic segmentation algorithm is 5 frames/s, the average accuracy for three classes of objects according to the IoU metric is 92%.
Author Perevoshchikova, A.A.
Medvedeva, E.V.
Author_xml – sequence: 1
  givenname: E.V.
  surname: Medvedeva
  fullname: Medvedeva, E.V.
– sequence: 2
  givenname: A.A.
  surname: Perevoshchikova
  fullname: Perevoshchikova, A.A.
BookMark eNo9kNtqwzAMhs3oYF3XF9iVXyCbD0mcXI6wQ6FQGNu1cRU5dZfGw05Xuqdf0o5eSIJf6BN8t2TS-Q4JuefsgReiUI8i5SLJuSqTapXwLJdXZHrJJmTKuEwTkWbihsxj3DLGhqucp3xK6neMaAJsqLe0w30w7TD6gw9f1LSND67f7CK1PtCA4JvO_bquocG49mCO1HU2mNiHPfT7gNSvtwh9HGL642r01O1Mg_GOXFvTRpz_zxn5fHn-qN6S5ep1UT0tExCy6BOhSsihQFYIULwAgcyua45jSzNWGyksIgjDcyVzxeSwFiUbCgRkZS1nZHHm1t5s9XcYvoej9sbpU-BDo03oHbSoc1CZKXnNCyXTUnIjbbkGloEUBaapGVjizILgYwxoLzzO9Em7HhXrUbGuVnrULv8AQFl48g
Cites_doi 10.1016/j.hspr.2023.01.001
10.1109/AERO53065.2022.9843537
10.30932/9785002182794-2023-620-625
10.21046/2070-7401-2021-18-6-35-45
10.1109/ICSP58490.2023.10248526
10.1016/j.autcon.2023.105069
10.23919/FRUCT50888.2021.9347653
10.14498/tech.2022.1.4
10.1016/j.measurement.2022.111277
10.18196/26123
10.1109/ICCEAI52939.2021.00075
10.1109/ICITCS.2013.6717896
10.1016/j.jii.2024.100571
10.1109/TENCON58879.2023.10322378
10.1016/j.procs.2022.12.031
10.1007/978-3-319-46448-0_2
10.30932/1992-3252-2019-17-62-72
10.1109/ICCV.2017.322
10.1109/NNICE58320.2023.10105805
10.1007/978-3-319-11430-9_11
10.1109/ICAAIC56838.2023.10140366
ContentType Journal Article
DBID AAYXX
CITATION
DOA
DOI 10.18287/2412-6179-CO-1563
DatabaseName CrossRef
DOAJ Directory of Open Access Journals
DatabaseTitle CrossRef
DatabaseTitleList CrossRef
Database_xml – sequence: 1
  dbid: DOA
  name: DOAJ Directory of Open Access Journals
  url: https://www.doaj.org/
  sourceTypes: Open Website
DeliveryMethod fulltext_linktorsrc
Discipline Applied Sciences
EISSN 2412-6179
EndPage 450
ExternalDocumentID oai_doaj_org_article_6c75a91d18734931a3f9bc05c328e44a
10_18287_2412_6179_CO_1563
GroupedDBID 642
AAFWJ
AAYXX
ADBBV
AFPKN
ALMA_UNASSIGNED_HOLDINGS
BCNDV
CITATION
GROUPED_DOAJ
ID FETCH-LOGICAL-c238t-279c6c8e082c718c2e0fbd1efbd1450da32feec2a167367030fb290b29c2c59d3
IEDL.DBID DOA
ISICitedReferencesCount 0
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=001466104000009&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 0134-2452
IngestDate Fri Oct 03 12:44:31 EDT 2025
Sat Oct 25 05:10:23 EDT 2025
IsDoiOpenAccess true
IsOpenAccess true
IsPeerReviewed true
IsScholarly true
Issue 3
Language English
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-c238t-279c6c8e082c718c2e0fbd1efbd1450da32feec2a167367030fb290b29c2c59d3
OpenAccessLink https://doaj.org/article/6c75a91d18734931a3f9bc05c328e44a
PageCount 8
ParticipantIDs doaj_primary_oai_doaj_org_article_6c75a91d18734931a3f9bc05c328e44a
crossref_primary_10_18287_2412_6179_CO_1563
PublicationCentury 2000
PublicationDate 2025-06-01
PublicationDateYYYYMMDD 2025-06-01
PublicationDate_xml – month: 06
  year: 2025
  text: 2025-06-01
  day: 01
PublicationDecade 2020
PublicationTitle Kompʹûternaâ optika
PublicationYear 2025
Publisher Samara National Research University
Publisher_xml – name: Samara National Research University
References ref13
ref12
ref15
ref14
ref20
ref11
ref22
ref10
ref21
ref0
ref2
ref1
ref17
ref16
ref19
ref18
ref8
ref7
ref9
ref4
ref3
ref6
ref5
References_xml – ident: ref12
  doi: 10.1016/j.hspr.2023.01.001
– ident: ref13
  doi: 10.1109/AERO53065.2022.9843537
– ident: ref1
  doi: 10.30932/9785002182794-2023-620-625
– ident: ref6
  doi: 10.21046/2070-7401-2021-18-6-35-45
– ident: ref18
  doi: 10.1109/ICSP58490.2023.10248526
– ident: ref11
  doi: 10.1016/j.autcon.2023.105069
– ident: ref7
  doi: 10.23919/FRUCT50888.2021.9347653
– ident: ref8
  doi: 10.14498/tech.2022.1.4
– ident: ref3
  doi: 10.1016/j.measurement.2022.111277
– ident: ref14
  doi: 10.18196/26123
– ident: ref16
  doi: 10.1109/ICCEAI52939.2021.00075
– ident: ref5
  doi: 10.1109/ICITCS.2013.6717896
– ident: ref2
  doi: 10.1016/j.jii.2024.100571
– ident: ref20
  doi: 10.1109/TENCON58879.2023.10322378
– ident: ref10
  doi: 10.1016/j.procs.2022.12.031
– ident: ref15
  doi: 10.1007/978-3-319-46448-0_2
– ident: ref9
  doi: 10.30932/1992-3252-2019-17-62-72
– ident: ref19
  doi: 10.1109/ICCV.2017.322
– ident: ref22
  doi: 10.1109/NNICE58320.2023.10105805
– ident: ref4
  doi: 10.1007/978-3-319-11430-9_11
– ident: ref21
  doi: 10.1109/ICAAIC56838.2023.10140366
– ident: ref0
– ident: ref17
SSID ssj0002876141
Score 2.3271546
Snippet The article describes the development of two neural network algorithms for recognizing objects of the railway infrastructure in video images. Both algorithms...
SourceID doaj
crossref
SourceType Open Website
Index Database
StartPage 443
SubjectTerms machine vision systems
neural network algorithms
object detection
railway infrastructure objects
railway traffic safety
semantic segmentation
Title Research of neural network algorithms for recognizing railway infrastructure objects in video images
URI https://doaj.org/article/6c75a91d18734931a3f9bc05c328e44a
Volume 49
WOSCitedRecordID wos001466104000009&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText 1
inHoldings 1
isFullTextHit
isPrint
journalDatabaseRights – providerCode: PRVAON
  databaseName: DOAJ Directory of Open Access Journals
  customDbUrl:
  eissn: 2412-6179
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0002876141
  issn: 0134-2452
  databaseCode: DOA
  dateStart: 19870101
  isFulltext: true
  titleUrlDefault: https://www.doaj.org/
  providerName: Directory of Open Access Journals
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwrV07T8MwELZQxcDCG1Fe8sCGrDq28_AIFRVTywBSt8h2bChqG9QUEPx67pIUlYmFIRmsJIo-P747--47Qi5lGmxqwTuJg-VMSWWYFRlnqQmwXMa8cNzWxSbS4TAbj_X9WqkvjAlr5IEb4HqJS2OjoyLKUqm0jIwM2joeOykyr1RtGvFUrzlTL_WWEbjnqilGKBXD48U2YwYF3ntAWwKT4zTrjxh4MPIXK62J99csM9gl2615SK-b39ojG36-T3ZaU5G2E7E6IMUqYo6WgaImJbw0byK6qZk-leDyP88qChYpbUOEvoCj6MJMph_mk8KwWphGOvZt4WlpcTemgmaKaXklncxgmakOyePg9qF_x9qCCcwB8y6ZSLVLXOaB1h1wjhOeB1tEHm8KcDdSBO-dMBFGc-FcD1ZoDpcTLtaFPCKdeTn3x4SqzMbeG8md10p4kdkkyYQO3EgrvEu65GoFWP7a6GLk6E8gvDnCmyO8eX-UI7xdcoOY_jyJmtZ1A_R03vZ0_ldPn_zHR07JlsAKvvU-yhnpANL-nGy69-WkWlzUg-gbxCXJ0A
linkProvider Directory of Open Access Journals
openUrl ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Research+of+neural+network+algorithms+for+recognizing+railway+infrastructure+objects+in+video+images&rft.jtitle=Komp%CA%B9%C3%BBterna%C3%A2+optika&rft.au=E.V.+Medvedeva&rft.au=A.A.+Perevoshchikova&rft.date=2025-06-01&rft.pub=Samara+National+Research+University&rft.issn=0134-2452&rft.eissn=2412-6179&rft.volume=49&rft.issue=3&rft.spage=443&rft.epage=450&rft_id=info:doi/10.18287%2F2412-6179-CO-1563&rft.externalDBID=DOA&rft.externalDocID=oai_doaj_org_article_6c75a91d18734931a3f9bc05c328e44a
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=0134-2452&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=0134-2452&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=0134-2452&client=summon