Research on Fabric Defect Detection Algorithm Based on Lightweight YOLOv7-Tiny

The current advanced neural network models are expanding in size and complexity to achieve improved detection accuracy. This study designs a lightweight fabric defect detection algorithm based on YOLOv7-tiny, called YOLOv7-tiny-MGCK. Its objectives are to improve the performance of fabric defect det...

Full description

Saved in:
Bibliographic Details
Main Authors: Tang Li, Mei Shunqi, Shi Yishan, Zhou Shi, Zheng Quan, Hongkai Jiang, Xu Qiao, Zhang Zhiming
Format: Article
Language:English
Published: Taylor & Francis Group 2024-12-01
Series:Journal of Natural Fibers
Subjects:
Online Access:https://www.tandfonline.com/doi/10.1080/15440478.2024.2352753
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850164301278478336
author Tang Li
Mei Shunqi
Shi Yishan
Zhou Shi
Zheng Quan
Hongkai Jiang
Xu Qiao
Zhang Zhiming
author_facet Tang Li
Mei Shunqi
Shi Yishan
Zhou Shi
Zheng Quan
Hongkai Jiang
Xu Qiao
Zhang Zhiming
author_sort Tang Li
collection DOAJ
description The current advanced neural network models are expanding in size and complexity to achieve improved detection accuracy. This study designs a lightweight fabric defect detection algorithm based on YOLOv7-tiny, called YOLOv7-tiny-MGCK. Its objectives are to improve the performance of fabric defect detection against complex backgrounds and to find a balance between the algorithm’s lightweight nature and its accuracy. The algorithm utilizes the Mish activation function, known for its superior nonlinear performance capability and smoother curve, enabling the neural network to manage more complex challenges. The Ghost convolution module is also incorporated to reduce computation and model parameters. The lightweight upsampling technique CARAFE facilitates the flexible extraction of deep features, coupled with their integration with shallow features. In addition, an improved K-Means clustering algorithm, KMMP, is employed to select appropriate anchor box for fabric defects. The experimental results show: a reduction in the number of parameters by 45.5% and computational volume by 41.0%, along with increases in precision by 3.9%, recall by 7.0%, and mAP by 3.0%. These results indicated that the improved algorithm achieves a more effective balance between detection performance and the requirement for a lightweight solution.
format Article
id doaj-art-a4cd021ce1274015bb3affcbe29ef138
institution OA Journals
issn 1544-0478
1544-046X
language English
publishDate 2024-12-01
publisher Taylor & Francis Group
record_format Article
series Journal of Natural Fibers
spelling doaj-art-a4cd021ce1274015bb3affcbe29ef1382025-08-20T02:22:02ZengTaylor & Francis GroupJournal of Natural Fibers1544-04781544-046X2024-12-0121110.1080/15440478.2024.2352753Research on Fabric Defect Detection Algorithm Based on Lightweight YOLOv7-TinyTang Li0Mei Shunqi1Shi Yishan2Zhou Shi3Zheng Quan4Hongkai Jiang5Xu Qiao6Zhang Zhiming7Hubei Digital Textile Equipment Key Laboratory, Wuhan Textile University, Wuhan, ChinaHubei Digital Textile Equipment Key Laboratory, Wuhan Textile University, Wuhan, ChinaHubei Digital Textile Equipment Key Laboratory, Wuhan Textile University, Wuhan, ChinaHubei Digital Textile Equipment Key Laboratory, Wuhan Textile University, Wuhan, ChinaHubei Digital Textile Equipment Key Laboratory, Wuhan Textile University, Wuhan, ChinaThe Chinese University of Hongkong, Shenzhen, ChinaHubei Digital Textile Equipment Key Laboratory, Wuhan Textile University, Wuhan, ChinaHubei Digital Textile Equipment Key Laboratory, Wuhan Textile University, Wuhan, ChinaThe current advanced neural network models are expanding in size and complexity to achieve improved detection accuracy. This study designs a lightweight fabric defect detection algorithm based on YOLOv7-tiny, called YOLOv7-tiny-MGCK. Its objectives are to improve the performance of fabric defect detection against complex backgrounds and to find a balance between the algorithm’s lightweight nature and its accuracy. The algorithm utilizes the Mish activation function, known for its superior nonlinear performance capability and smoother curve, enabling the neural network to manage more complex challenges. The Ghost convolution module is also incorporated to reduce computation and model parameters. The lightweight upsampling technique CARAFE facilitates the flexible extraction of deep features, coupled with their integration with shallow features. In addition, an improved K-Means clustering algorithm, KMMP, is employed to select appropriate anchor box for fabric defects. The experimental results show: a reduction in the number of parameters by 45.5% and computational volume by 41.0%, along with increases in precision by 3.9%, recall by 7.0%, and mAP by 3.0%. These results indicated that the improved algorithm achieves a more effective balance between detection performance and the requirement for a lightweight solution.https://www.tandfonline.com/doi/10.1080/15440478.2024.2352753Neural networkfabric defect detectionactivation functionGhost moduleupsamplingclustering algorithm
spellingShingle Tang Li
Mei Shunqi
Shi Yishan
Zhou Shi
Zheng Quan
Hongkai Jiang
Xu Qiao
Zhang Zhiming
Research on Fabric Defect Detection Algorithm Based on Lightweight YOLOv7-Tiny
Journal of Natural Fibers
Neural network
fabric defect detection
activation function
Ghost module
upsampling
clustering algorithm
title Research on Fabric Defect Detection Algorithm Based on Lightweight YOLOv7-Tiny
title_full Research on Fabric Defect Detection Algorithm Based on Lightweight YOLOv7-Tiny
title_fullStr Research on Fabric Defect Detection Algorithm Based on Lightweight YOLOv7-Tiny
title_full_unstemmed Research on Fabric Defect Detection Algorithm Based on Lightweight YOLOv7-Tiny
title_short Research on Fabric Defect Detection Algorithm Based on Lightweight YOLOv7-Tiny
title_sort research on fabric defect detection algorithm based on lightweight yolov7 tiny
topic Neural network
fabric defect detection
activation function
Ghost module
upsampling
clustering algorithm
url https://www.tandfonline.com/doi/10.1080/15440478.2024.2352753
work_keys_str_mv AT tangli researchonfabricdefectdetectionalgorithmbasedonlightweightyolov7tiny
AT meishunqi researchonfabricdefectdetectionalgorithmbasedonlightweightyolov7tiny
AT shiyishan researchonfabricdefectdetectionalgorithmbasedonlightweightyolov7tiny
AT zhoushi researchonfabricdefectdetectionalgorithmbasedonlightweightyolov7tiny
AT zhengquan researchonfabricdefectdetectionalgorithmbasedonlightweightyolov7tiny
AT hongkaijiang researchonfabricdefectdetectionalgorithmbasedonlightweightyolov7tiny
AT xuqiao researchonfabricdefectdetectionalgorithmbasedonlightweightyolov7tiny
AT zhangzhiming researchonfabricdefectdetectionalgorithmbasedonlightweightyolov7tiny