LightYOLO: Lightweight model based on YOLOv8n for defect detection of ultrasonically welded wire terminations

Defect inspection of the surface in ultrasonically welded wire terminations is an important inspection procedure to ensure welding quality. However, the detection task of ultrasonic welding defects based on deep learning still faces the challenges of low detection accuracy and slow inference speed....

Full description

Saved in:
Bibliographic Details
Main Authors: Jianshu Xu, Lun Zhao, Yu Ren, Zhigang Li, Zeshan Abbas, Lan Zhang, Md Shafiqul Islam
Format: Article
Language:English
Published: Elsevier 2024-12-01
Series:Engineering Science and Technology, an International Journal
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S2215098624002829
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850112098401517568
author Jianshu Xu
Lun Zhao
Yu Ren
Zhigang Li
Zeshan Abbas
Lan Zhang
Md Shafiqul Islam
author_facet Jianshu Xu
Lun Zhao
Yu Ren
Zhigang Li
Zeshan Abbas
Lan Zhang
Md Shafiqul Islam
author_sort Jianshu Xu
collection DOAJ
description Defect inspection of the surface in ultrasonically welded wire terminations is an important inspection procedure to ensure welding quality. However, the detection task of ultrasonic welding defects based on deep learning still faces the challenges of low detection accuracy and slow inference speed. Therefore, to solve the above problems, we propose a fast and effective lightweight detection model based on You Only Look Once v8 (YOLOv8n), named LightYOLO. Specifically, first, to achieve fast feature extraction, a Two-Convolution module with FasterNet block and Efficient multi-scale attention (CTFE) structures is introduced in the backbone network. Secondly, Group-Shuffle Convolution (GSConv) is used to construct the feature fusion structure of the neck, which enhances the fusion efficiency of multi-level features. Finally, an auxiliary head training method is introduced to extract shallow details of the network. To verify the effectiveness of the proposed method, we constructed a surface defect data set of ultrasonic welding wire terminals and conducted a series of experiments. The results of experiments show that the precision of LightYOLO is 93.4%, which is 3.5% higher than YOLOv8n(89.9%). In addition, the model size was reduced to 1/2 of the baseline model. LightYOLO shows the potential for rapid detection on edge computing devices. The source code and dataset for our project is accessible at https://github.com/JianshuXu/LightYOLO.
format Article
id doaj-art-62691ef5cf394f39bb835226a96bee93
institution OA Journals
issn 2215-0986
language English
publishDate 2024-12-01
publisher Elsevier
record_format Article
series Engineering Science and Technology, an International Journal
spelling doaj-art-62691ef5cf394f39bb835226a96bee932025-08-20T02:37:29ZengElsevierEngineering Science and Technology, an International Journal2215-09862024-12-016010189610.1016/j.jestch.2024.101896LightYOLO: Lightweight model based on YOLOv8n for defect detection of ultrasonically welded wire terminationsJianshu Xu0Lun Zhao1Yu Ren2Zhigang Li3Zeshan Abbas4Lan Zhang5Md Shafiqul Islam6School of Mechanical and Electrical Engineering, Yunnan Open University, Kunming, 650500, China; School of Electronic and Information Engineering, University of Science and Technology Liaoning, Anshan, 114051, ChinaSchool of Mechanical and Electrical Engineering, Yunnan Open University, Kunming, 650500, China; Department of Mechanical Engineering, Blekinge Institute of Technology, Karlskrona, 37179, Sweden; Corresponding author at: School of Mechanical and Electrical Engineering, Yunnan Open University, Kunming, 650500, China.School of Mechanical and Electrical Engineering, Yunnan Open University, Kunming, 650500, China; School of Biomedical Engineering, Shenzhen University, Shenzhen, 518060, ChinaSchool of Electronic and Information Engineering, University of Science and Technology Liaoning, Anshan, 114051, ChinaSchool of Mechanical and Electrical Engineering, Yunnan Open University, Kunming, 650500, ChinaSchool of Mechanical and Electrical Engineering, Yunnan Open University, Kunming, 650500, ChinaDepartment of Mechanical Engineering, Blekinge Institute of Technology, Karlskrona, 37179, SwedenDefect inspection of the surface in ultrasonically welded wire terminations is an important inspection procedure to ensure welding quality. However, the detection task of ultrasonic welding defects based on deep learning still faces the challenges of low detection accuracy and slow inference speed. Therefore, to solve the above problems, we propose a fast and effective lightweight detection model based on You Only Look Once v8 (YOLOv8n), named LightYOLO. Specifically, first, to achieve fast feature extraction, a Two-Convolution module with FasterNet block and Efficient multi-scale attention (CTFE) structures is introduced in the backbone network. Secondly, Group-Shuffle Convolution (GSConv) is used to construct the feature fusion structure of the neck, which enhances the fusion efficiency of multi-level features. Finally, an auxiliary head training method is introduced to extract shallow details of the network. To verify the effectiveness of the proposed method, we constructed a surface defect data set of ultrasonic welding wire terminals and conducted a series of experiments. The results of experiments show that the precision of LightYOLO is 93.4%, which is 3.5% higher than YOLOv8n(89.9%). In addition, the model size was reduced to 1/2 of the baseline model. LightYOLO shows the potential for rapid detection on edge computing devices. The source code and dataset for our project is accessible at https://github.com/JianshuXu/LightYOLO.http://www.sciencedirect.com/science/article/pii/S2215098624002829Ultrasonic metal weldingDeep learningObject detectionLightweight
spellingShingle Jianshu Xu
Lun Zhao
Yu Ren
Zhigang Li
Zeshan Abbas
Lan Zhang
Md Shafiqul Islam
LightYOLO: Lightweight model based on YOLOv8n for defect detection of ultrasonically welded wire terminations
Engineering Science and Technology, an International Journal
Ultrasonic metal welding
Deep learning
Object detection
Lightweight
title LightYOLO: Lightweight model based on YOLOv8n for defect detection of ultrasonically welded wire terminations
title_full LightYOLO: Lightweight model based on YOLOv8n for defect detection of ultrasonically welded wire terminations
title_fullStr LightYOLO: Lightweight model based on YOLOv8n for defect detection of ultrasonically welded wire terminations
title_full_unstemmed LightYOLO: Lightweight model based on YOLOv8n for defect detection of ultrasonically welded wire terminations
title_short LightYOLO: Lightweight model based on YOLOv8n for defect detection of ultrasonically welded wire terminations
title_sort lightyolo lightweight model based on yolov8n for defect detection of ultrasonically welded wire terminations
topic Ultrasonic metal welding
Deep learning
Object detection
Lightweight
url http://www.sciencedirect.com/science/article/pii/S2215098624002829
work_keys_str_mv AT jianshuxu lightyololightweightmodelbasedonyolov8nfordefectdetectionofultrasonicallyweldedwireterminations
AT lunzhao lightyololightweightmodelbasedonyolov8nfordefectdetectionofultrasonicallyweldedwireterminations
AT yuren lightyololightweightmodelbasedonyolov8nfordefectdetectionofultrasonicallyweldedwireterminations
AT zhigangli lightyololightweightmodelbasedonyolov8nfordefectdetectionofultrasonicallyweldedwireterminations
AT zeshanabbas lightyololightweightmodelbasedonyolov8nfordefectdetectionofultrasonicallyweldedwireterminations
AT lanzhang lightyololightweightmodelbasedonyolov8nfordefectdetectionofultrasonicallyweldedwireterminations
AT mdshafiqulislam lightyololightweightmodelbasedonyolov8nfordefectdetectionofultrasonicallyweldedwireterminations