Skip to main navigation Skip to search Skip to main content

Crack identification inside on-site steel box girder based on fusion convolutional neural network

  • Ministry of Industry and Information Technology
  • Ministry of Education of the People's Republic of China
  • School of Civil Engineering, Harbin Institute of Technology

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

In this paper we propose a novel fusion convolutional neural network to identify the local fatigue cracks in steel box girder of cable-stayed bridge. Unlike conventional CNN's chain-like structure, the proposed network fully exploits multiscale and multilevel information of input images by combining all the meaningful convolutional features together. Raw images with high resolution of 3624×4928 are decomposed into three kinds of sub-image sets with lower resolution of 64×64, background, handwriting and crack, respectively. Multi-functional layers are stacked including convolution, ReLU, softmaxResults show that the test error drops to 4% after only 50 epochs and it is more effective compared with other deep learning networks when handling large image datasets.

Original languageEnglish
Title of host publicationSensors and Smart Structures Technologies for Civil, Mechanical, and Aerospace Systems 2018
EditorsKon-Well Wang, Hoon Sohn, Jerome P. Lynch
PublisherSPIE
ISBN (Electronic)9781510616929
DOIs
StatePublished - 2018
Externally publishedYes
EventSensors and Smart Structures Technologies for Civil, Mechanical, and Aerospace Systems 2018 - Denver, United States
Duration: 5 Mar 20188 Mar 2018

Publication series

NameProceedings of SPIE - The International Society for Optical Engineering
Volume10598
ISSN (Print)0277-786X
ISSN (Electronic)1996-756X

Conference

ConferenceSensors and Smart Structures Technologies for Civil, Mechanical, and Aerospace Systems 2018
Country/TerritoryUnited States
CityDenver
Period5/03/188/03/18

Keywords

  • Crack identification
  • convolutional neural network
  • multiscale and multilevel features
  • steel box girder

Fingerprint

Dive into the research topics of 'Crack identification inside on-site steel box girder based on fusion convolutional neural network'. Together they form a unique fingerprint.

Cite this