The improved C3STR structure.

<div><p>Aiming at the problems of low accuracy and large computation in the task of PCB defect detection. This paper proposes a lightweight PCB defect detection algorithm based on YOLO. To address the problem of large numbers of parameters and calculations, GhostNet are used in Backbone...

وصف كامل

محفوظ في:
التفاصيل البيبلوغرافية
المؤلف الرئيسي: Shiwei Yu (6060308) (author)
مؤلفون آخرون: Feng Pan (257056) (author), Xiaoqiang Zhang (2413909) (author), Linhua Zhou (6430619) (author), Liang Zhang (41392) (author), Jikui Wang (4519225) (author)
منشور في: 2025
الموضوعات:
الوسوم: إضافة وسم
لا توجد وسوم, كن أول من يضع وسما على هذه التسجيلة!
_version_ 1852021197173161984
author Shiwei Yu (6060308)
author2 Feng Pan (257056)
Xiaoqiang Zhang (2413909)
Linhua Zhou (6430619)
Liang Zhang (41392)
Jikui Wang (4519225)
author2_role author
author
author
author
author
author_facet Shiwei Yu (6060308)
Feng Pan (257056)
Xiaoqiang Zhang (2413909)
Linhua Zhou (6430619)
Liang Zhang (41392)
Jikui Wang (4519225)
author_role author
dc.creator.none.fl_str_mv Shiwei Yu (6060308)
Feng Pan (257056)
Xiaoqiang Zhang (2413909)
Linhua Zhou (6430619)
Liang Zhang (41392)
Jikui Wang (4519225)
dc.date.none.fl_str_mv 2025-04-18T17:25:48Z
dc.identifier.none.fl_str_mv 10.1371/journal.pone.0320344.g005
dc.relation.none.fl_str_mv https://figshare.com/articles/figure/The_improved_C3STR_structure_/28825334
dc.rights.none.fl_str_mv CC BY 4.0
info:eu-repo/semantics/openAccess
dc.subject.none.fl_str_mv Space Science
Environmental Sciences not elsewhere classified
Biological Sciences not elsewhere classified
Mathematical Sciences not elsewhere classified
Information Systems not elsewhere classified
xlink "> aiming
simple defect types
pcb defect detection
lightweight detection algorithm
depthwise separable convolution
redundant parameters within
model &# 8217
yolo </ p
panet network structure
ordinary convolution
model lightweight
scale features
results indicated
parameter count
paper proposes
original model
large numbers
large computation
defective images
confusion caused
cluttered background
c3str module
c3 module
better suited
arithmetic platforms
dc.title.none.fl_str_mv The improved C3STR structure.
dc.type.none.fl_str_mv Image
Figure
info:eu-repo/semantics/publishedVersion
image
description <div><p>Aiming at the problems of low accuracy and large computation in the task of PCB defect detection. This paper proposes a lightweight PCB defect detection algorithm based on YOLO. To address the problem of large numbers of parameters and calculations, GhostNet are used in Backbone to keep the model lightweight. Second, the ordinary convolution of the neck network is improved by depthwise separable convolution, resulting in a reduction of redundant parameters within the neck network. Afterwards, the Swin-Transformer is integrated with the C3 module in the Neck to build the C3STR module, which aims to address the issue of cluttered background in defective images and the confusion caused by simple defect types. Finally, the PANet network structure is replaced with the bidirectional feature pyramid network (BIFPN) structure to enhance the fusion of multi-scale features in the network. The results indicated that when comparing our model with the original model, there was a 47.2% reduction in the model’s parameter count, a 48.5% reduction in GFLOPs, a 42.4% reduction in Weight, a 2.0% reduction in FPS, and a 2.4% rise in mAP. The model is better suited for use on low-arithmetic platforms as a result.</p></div>
eu_rights_str_mv openAccess
id Manara_bd0bbcd2bc9900b035dd39a08c6ab759
identifier_str_mv 10.1371/journal.pone.0320344.g005
network_acronym_str Manara
network_name_str ManaraRepo
oai_identifier_str oai:figshare.com:article/28825334
publishDate 2025
repository.mail.fl_str_mv
repository.name.fl_str_mv
repository_id_str
rights_invalid_str_mv CC BY 4.0
spelling The improved C3STR structure.Shiwei Yu (6060308)Feng Pan (257056)Xiaoqiang Zhang (2413909)Linhua Zhou (6430619)Liang Zhang (41392)Jikui Wang (4519225)Space ScienceEnvironmental Sciences not elsewhere classifiedBiological Sciences not elsewhere classifiedMathematical Sciences not elsewhere classifiedInformation Systems not elsewhere classifiedxlink "> aimingsimple defect typespcb defect detectionlightweight detection algorithmdepthwise separable convolutionredundant parameters withinmodel &# 8217yolo </ ppanet network structureordinary convolutionmodel lightweightscale featuresresults indicatedparameter countpaper proposesoriginal modellarge numberslarge computationdefective imagesconfusion causedcluttered backgroundc3str modulec3 modulebetter suitedarithmetic platforms<div><p>Aiming at the problems of low accuracy and large computation in the task of PCB defect detection. This paper proposes a lightweight PCB defect detection algorithm based on YOLO. To address the problem of large numbers of parameters and calculations, GhostNet are used in Backbone to keep the model lightweight. Second, the ordinary convolution of the neck network is improved by depthwise separable convolution, resulting in a reduction of redundant parameters within the neck network. Afterwards, the Swin-Transformer is integrated with the C3 module in the Neck to build the C3STR module, which aims to address the issue of cluttered background in defective images and the confusion caused by simple defect types. Finally, the PANet network structure is replaced with the bidirectional feature pyramid network (BIFPN) structure to enhance the fusion of multi-scale features in the network. The results indicated that when comparing our model with the original model, there was a 47.2% reduction in the model’s parameter count, a 48.5% reduction in GFLOPs, a 42.4% reduction in Weight, a 2.0% reduction in FPS, and a 2.4% rise in mAP. The model is better suited for use on low-arithmetic platforms as a result.</p></div>2025-04-18T17:25:48ZImageFigureinfo:eu-repo/semantics/publishedVersionimage10.1371/journal.pone.0320344.g005https://figshare.com/articles/figure/The_improved_C3STR_structure_/28825334CC BY 4.0info:eu-repo/semantics/openAccessoai:figshare.com:article/288253342025-04-18T17:25:48Z
spellingShingle The improved C3STR structure.
Shiwei Yu (6060308)
Space Science
Environmental Sciences not elsewhere classified
Biological Sciences not elsewhere classified
Mathematical Sciences not elsewhere classified
Information Systems not elsewhere classified
xlink "> aiming
simple defect types
pcb defect detection
lightweight detection algorithm
depthwise separable convolution
redundant parameters within
model &# 8217
yolo </ p
panet network structure
ordinary convolution
model lightweight
scale features
results indicated
parameter count
paper proposes
original model
large numbers
large computation
defective images
confusion caused
cluttered background
c3str module
c3 module
better suited
arithmetic platforms
status_str publishedVersion
title The improved C3STR structure.
title_full The improved C3STR structure.
title_fullStr The improved C3STR structure.
title_full_unstemmed The improved C3STR structure.
title_short The improved C3STR structure.
title_sort The improved C3STR structure.
topic Space Science
Environmental Sciences not elsewhere classified
Biological Sciences not elsewhere classified
Mathematical Sciences not elsewhere classified
Information Systems not elsewhere classified
xlink "> aiming
simple defect types
pcb defect detection
lightweight detection algorithm
depthwise separable convolution
redundant parameters within
model &# 8217
yolo </ p
panet network structure
ordinary convolution
model lightweight
scale features
results indicated
parameter count
paper proposes
original model
large numbers
large computation
defective images
confusion caused
cluttered background
c3str module
c3 module
better suited
arithmetic platforms