Overall structure of the model.

<div><p>Drum roller surface defect detection is of great research significance for control production quality. Aiming at solving the problems that the traditional light source visual imaging system, which does not clearly reflect defect features, the defect detection efficiency is low, a...

Full description

Saved in:
Bibliographic Details
Main Author: Guofeng Qin (11025170) (author)
Other Authors: Qinkai Zou (20678485) (author), Mengyan Li (1858759) (author), Yi Deng (421971) (author), Peiwen Mi (20678488) (author), Yongjian Zhu (6562817) (author), Hao Liu (12832) (author)
Published: 2025
Subjects:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1852022999256924160
author Guofeng Qin (11025170)
author2 Qinkai Zou (20678485)
Mengyan Li (1858759)
Yi Deng (421971)
Peiwen Mi (20678488)
Yongjian Zhu (6562817)
Hao Liu (12832)
author2_role author
author
author
author
author
author
author_facet Guofeng Qin (11025170)
Qinkai Zou (20678485)
Mengyan Li (1858759)
Yi Deng (421971)
Peiwen Mi (20678488)
Yongjian Zhu (6562817)
Hao Liu (12832)
author_role author
dc.creator.none.fl_str_mv Guofeng Qin (11025170)
Qinkai Zou (20678485)
Mengyan Li (1858759)
Yi Deng (421971)
Peiwen Mi (20678488)
Yongjian Zhu (6562817)
Hao Liu (12832)
dc.date.none.fl_str_mv 2025-02-05T18:32:13Z
dc.identifier.none.fl_str_mv 10.1371/journal.pone.0316569.g007
dc.relation.none.fl_str_mv https://figshare.com/articles/figure/Overall_structure_of_the_model_/28353919
dc.rights.none.fl_str_mv CC BY 4.0
info:eu-repo/semantics/openAccess
dc.subject.none.fl_str_mv Science Policy
Space Science
Biological Sciences not elsewhere classified
Information Systems not elsewhere classified
model &# 8217
great research significance
feature extraction ability
experimental results show
broader dataset diversity
traditional light source
achieve efficient detection
using enhanced yolov8n
defect detection efficiency
main average accuracy
industrial drum rollers
improved yolov8n algorithm
control production quality
surface defect detection
system light source
production quality
drum rollers
structured light
defect characteristics
detection time
detection accuracy
rolling surface
algorithm designed
quality samples
drum roller
standard convolution
replace ciou
network pays
loss function
image acquisition
fusion network
certain improvement
better ensure
backbone network
2 %,
dc.title.none.fl_str_mv Overall structure of the model.
dc.type.none.fl_str_mv Image
Figure
info:eu-repo/semantics/publishedVersion
image
description <div><p>Drum roller surface defect detection is of great research significance for control production quality. Aiming at solving the problems that the traditional light source visual imaging system, which does not clearly reflect defect features, the defect detection efficiency is low, and the accuracy is not enough, this paper designs an image acquisition system based on line fringe structured light and proposes an improved deep learning network model based on YOLOv8n to achieve efficient detection of defects on the rolling surface of a drum roller. In the aspect of image acquisition, this paper selected the line fringe structured light as the system light source, which made up for the problem that the traditional light source does not reflect the defect characteristics. In terms of algorithms, firstly, using deformable convolution instead of standard convolution to enhance the feature extraction ability of the backbone network. Then, a new feature fusion module was proposed to enable the fusion network to learn additional original information. Finally, Wise-IoU was applied to replace CIoU in the loss function, so that the network pays more attention to the high-quality samples. The experimental results show that the improved YOLOv8n algorithm has a certain improvement in detection accuracy. The main average accuracy (mAP) is 97.2%, and the detection time is 4.3ms. The system and algorithm designed in this paper can better ensure the production quality of drum rollers. While effective, the model’s standard rectangular bounding boxes may limit precision for elongated defects. Future work could explore rotated bounding boxes and broader dataset diversity to enhance generalization in real-world applications.</p></div>
eu_rights_str_mv openAccess
id Manara_4c4161f65fca595bd7219c9d1f7ff080
identifier_str_mv 10.1371/journal.pone.0316569.g007
network_acronym_str Manara
network_name_str ManaraRepo
oai_identifier_str oai:figshare.com:article/28353919
publishDate 2025
repository.mail.fl_str_mv
repository.name.fl_str_mv
repository_id_str
rights_invalid_str_mv CC BY 4.0
spelling Overall structure of the model.Guofeng Qin (11025170)Qinkai Zou (20678485)Mengyan Li (1858759)Yi Deng (421971)Peiwen Mi (20678488)Yongjian Zhu (6562817)Hao Liu (12832)Science PolicySpace ScienceBiological Sciences not elsewhere classifiedInformation Systems not elsewhere classifiedmodel &# 8217great research significancefeature extraction abilityexperimental results showbroader dataset diversitytraditional light sourceachieve efficient detectionusing enhanced yolov8ndefect detection efficiencymain average accuracyindustrial drum rollersimproved yolov8n algorithmcontrol production qualitysurface defect detectionsystem light sourceproduction qualitydrum rollersstructured lightdefect characteristicsdetection timedetection accuracyrolling surfacealgorithm designedquality samplesdrum rollerstandard convolutionreplace ciounetwork paysloss functionimage acquisitionfusion networkcertain improvementbetter ensurebackbone network2 %,<div><p>Drum roller surface defect detection is of great research significance for control production quality. Aiming at solving the problems that the traditional light source visual imaging system, which does not clearly reflect defect features, the defect detection efficiency is low, and the accuracy is not enough, this paper designs an image acquisition system based on line fringe structured light and proposes an improved deep learning network model based on YOLOv8n to achieve efficient detection of defects on the rolling surface of a drum roller. In the aspect of image acquisition, this paper selected the line fringe structured light as the system light source, which made up for the problem that the traditional light source does not reflect the defect characteristics. In terms of algorithms, firstly, using deformable convolution instead of standard convolution to enhance the feature extraction ability of the backbone network. Then, a new feature fusion module was proposed to enable the fusion network to learn additional original information. Finally, Wise-IoU was applied to replace CIoU in the loss function, so that the network pays more attention to the high-quality samples. The experimental results show that the improved YOLOv8n algorithm has a certain improvement in detection accuracy. The main average accuracy (mAP) is 97.2%, and the detection time is 4.3ms. The system and algorithm designed in this paper can better ensure the production quality of drum rollers. While effective, the model’s standard rectangular bounding boxes may limit precision for elongated defects. Future work could explore rotated bounding boxes and broader dataset diversity to enhance generalization in real-world applications.</p></div>2025-02-05T18:32:13ZImageFigureinfo:eu-repo/semantics/publishedVersionimage10.1371/journal.pone.0316569.g007https://figshare.com/articles/figure/Overall_structure_of_the_model_/28353919CC BY 4.0info:eu-repo/semantics/openAccessoai:figshare.com:article/283539192025-02-05T18:32:13Z
spellingShingle Overall structure of the model.
Guofeng Qin (11025170)
Science Policy
Space Science
Biological Sciences not elsewhere classified
Information Systems not elsewhere classified
model &# 8217
great research significance
feature extraction ability
experimental results show
broader dataset diversity
traditional light source
achieve efficient detection
using enhanced yolov8n
defect detection efficiency
main average accuracy
industrial drum rollers
improved yolov8n algorithm
control production quality
surface defect detection
system light source
production quality
drum rollers
structured light
defect characteristics
detection time
detection accuracy
rolling surface
algorithm designed
quality samples
drum roller
standard convolution
replace ciou
network pays
loss function
image acquisition
fusion network
certain improvement
better ensure
backbone network
2 %,
status_str publishedVersion
title Overall structure of the model.
title_full Overall structure of the model.
title_fullStr Overall structure of the model.
title_full_unstemmed Overall structure of the model.
title_short Overall structure of the model.
title_sort Overall structure of the model.
topic Science Policy
Space Science
Biological Sciences not elsewhere classified
Information Systems not elsewhere classified
model &# 8217
great research significance
feature extraction ability
experimental results show
broader dataset diversity
traditional light source
achieve efficient detection
using enhanced yolov8n
defect detection efficiency
main average accuracy
industrial drum rollers
improved yolov8n algorithm
control production quality
surface defect detection
system light source
production quality
drum rollers
structured light
defect characteristics
detection time
detection accuracy
rolling surface
algorithm designed
quality samples
drum roller
standard convolution
replace ciou
network pays
loss function
image acquisition
fusion network
certain improvement
better ensure
backbone network
2 %,