Computer-Aided Recognition of Defects in Welded Joints during Visual Inspections Based on Geometric Attributes

Abstract

Abstract: An automated defect recognition algorithm is presented for detecting and classifying weld defects by photographic images. The proposed recognition algorithm selects a defective domain in a segmented image, extracts geometric features from the image, and relates the defect to one of six classes: no defect, cavity, longitudinal crack, transverse crack, burn-through, or multiple defect. The algorithm is implemented in the Matlab 2018b MathWorks environment and has been tested on 60 photographs of defects of various classes; the accuracy of recognition was 85%.

Original language	English
Pages (from-to)	259-267
Number of pages	9
Journal	Russian Journal of Nondestructive Testing
Volume	56
Issue number	3
DOIs	https://doi.org/10.1134/S1061830920030055
Publication status	Published - 1 Mar 2020

Keywords

classification
defect
image processing
segmentation
visual inspection
weld

ASJC Scopus subject areas

Materials Science(all)
Condensed Matter Physics
Mechanics of Materials
Mechanical Engineering

Access to Document

10.1134/S1061830920030055

Fingerprint Dive into the research topics of 'Computer-Aided Recognition of Defects in Welded Joints during Visual Inspections Based on Geometric Attributes'. Together they form a unique fingerprint.

Cite this

Muravyov, S. V., & Pogadaeva, E. Y. (2020). Computer-Aided Recognition of Defects in Welded Joints during Visual Inspections Based on Geometric Attributes. Russian Journal of Nondestructive Testing, 56(3), 259-267. https://doi.org/10.1134/S1061830920030055

@article{cb8aa279715a48fb88a0a91178841334,

title = "Computer-Aided Recognition of Defects in Welded Joints during Visual Inspections Based on Geometric Attributes",

abstract = "Abstract: An automated defect recognition algorithm is presented for detecting and classifying weld defects by photographic images. The proposed recognition algorithm selects a defective domain in a segmented image, extracts geometric features from the image, and relates the defect to one of six classes: no defect, cavity, longitudinal crack, transverse crack, burn-through, or multiple defect. The algorithm is implemented in the Matlab 2018b MathWorks environment and has been tested on 60 photographs of defects of various classes; the accuracy of recognition was 85%.",

keywords = "classification, defect, image processing, segmentation, visual inspection, weld",

author = "Muravyov, {S. V.} and Pogadaeva, {E. Yu}",

year = "2020",

month = mar,

day = "1",

doi = "10.1134/S1061830920030055",

language = "English",

volume = "56",

pages = "259--267",

journal = "Russian Journal of Nondestructive Testing",

issn = "1061-8309",

publisher = "Maik Nauka-Interperiodica Publishing",

number = "3",

}

TY - JOUR

T1 - Computer-Aided Recognition of Defects in Welded Joints during Visual Inspections Based on Geometric Attributes

AU - Muravyov, S. V.

AU - Pogadaeva, E. Yu

PY - 2020/3/1

Y1 - 2020/3/1

N2 - Abstract: An automated defect recognition algorithm is presented for detecting and classifying weld defects by photographic images. The proposed recognition algorithm selects a defective domain in a segmented image, extracts geometric features from the image, and relates the defect to one of six classes: no defect, cavity, longitudinal crack, transverse crack, burn-through, or multiple defect. The algorithm is implemented in the Matlab 2018b MathWorks environment and has been tested on 60 photographs of defects of various classes; the accuracy of recognition was 85%.

AB - Abstract: An automated defect recognition algorithm is presented for detecting and classifying weld defects by photographic images. The proposed recognition algorithm selects a defective domain in a segmented image, extracts geometric features from the image, and relates the defect to one of six classes: no defect, cavity, longitudinal crack, transverse crack, burn-through, or multiple defect. The algorithm is implemented in the Matlab 2018b MathWorks environment and has been tested on 60 photographs of defects of various classes; the accuracy of recognition was 85%.

KW - classification

KW - defect

KW - image processing

KW - segmentation

KW - visual inspection

KW - weld

UR - http://www.scopus.com/inward/record.url?scp=85085986730&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85085986730&partnerID=8YFLogxK

U2 - 10.1134/S1061830920030055

DO - 10.1134/S1061830920030055

M3 - Article

AN - SCOPUS:85085986730

VL - 56

SP - 259

EP - 267

JO - Russian Journal of Nondestructive Testing

JF - Russian Journal of Nondestructive Testing

SN - 1061-8309

IS - 3

ER -