GSBYOLO: A lightweight Multi-Scale fusion network for road crack detection in complex environments

Abstract Timely detection and regular maintenance of road cracks are critical for road and traffic safety. However, existing detection methods face challenges such as varying target scales, large model parameters, and poor adaptability to complex backgrounds. To address these issues, this study prop...

Full description

Saved in:
Bibliographic Details
Main Authors: Yuhao Wang, Heran Zhu, Yirong Wang, Jianping Liu, Jun Xie, Bi Zhao, Siyue Zhao
Format: Article
Language:English
Published: Nature Portfolio 2025-07-01
Series:Scientific Reports
Subjects:
Online Access:https://doi.org/10.1038/s41598-025-11717-0
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849234684912861184
author Yuhao Wang
Heran Zhu
Yirong Wang
Jianping Liu
Jun Xie
Bi Zhao
Siyue Zhao
author_facet Yuhao Wang
Heran Zhu
Yirong Wang
Jianping Liu
Jun Xie
Bi Zhao
Siyue Zhao
author_sort Yuhao Wang
collection DOAJ
description Abstract Timely detection and regular maintenance of road cracks are critical for road and traffic safety. However, existing detection methods face challenges such as varying target scales, large model parameters, and poor adaptability to complex backgrounds. To address these issues, this study proposes an enhanced GSB-YOLO model. Inspired by the concepts of linear transformation and long-range attention mechanisms, a lightweight network structure was designed to reduce model parameters in the backbone network, thereby improving detection efficiency. Additionally, a novel SMC2f module was introduced in the neck structure, which calculates the “energy” of each neuron in the feature map, evaluates its contribution to the detection task, and dynamically assigns weighted coefficients. This method enhances the model’s detection robustness in complex backgrounds and effectively addresses the issue of insufficient emphasis on positive samples. Furthermore, through the optimization of the Path Aggregation Network (PAN) and the Bidirectional Feature Pyramid Network (BiFPN), efficient multi-scale feature fusion is achieved, further strengthening the model’s capacity to represent crack features at various scales. Experimental results indicate that the proposed GSB-YOLO model improves the mean average precision (mAP) in road crack detection tasks by 3.2%, demonstrating its significant application value in road crack detection and traffic safety assurance.
format Article
id doaj-art-59dfa280161044c2a7891c7a2677cecc
institution Kabale University
issn 2045-2322
language English
publishDate 2025-07-01
publisher Nature Portfolio
record_format Article
series Scientific Reports
spelling doaj-art-59dfa280161044c2a7891c7a2677cecc2025-08-20T04:03:03ZengNature PortfolioScientific Reports2045-23222025-07-0115111410.1038/s41598-025-11717-0GSBYOLO: A lightweight Multi-Scale fusion network for road crack detection in complex environmentsYuhao Wang0Heran Zhu1Yirong Wang2Jianping Liu3Jun Xie4Bi Zhao5Siyue Zhao6College of Water Conservancy and Hydropower, Sichuan Agricultural UniversityYunnan Tropical and Subtropical Animal Virus Disease Laboratory, Yunnan Academy of Animal Husbandry and Veterinary SciencesCollege of Water Conservancy and Hydropower, Sichuan Agricultural UniversityCollege of Water Conservancy and Hydropower, Sichuan Agricultural UniversityCollege of Water Conservancy and Hydropower, Sichuan Agricultural UniversityTea Research Institute, Yunnan Key Laboratory of Tea Science, Yunnan Academy of Agricultural SciencesCollege of Water Conservancy and Hydropower, Sichuan Agricultural UniversityAbstract Timely detection and regular maintenance of road cracks are critical for road and traffic safety. However, existing detection methods face challenges such as varying target scales, large model parameters, and poor adaptability to complex backgrounds. To address these issues, this study proposes an enhanced GSB-YOLO model. Inspired by the concepts of linear transformation and long-range attention mechanisms, a lightweight network structure was designed to reduce model parameters in the backbone network, thereby improving detection efficiency. Additionally, a novel SMC2f module was introduced in the neck structure, which calculates the “energy” of each neuron in the feature map, evaluates its contribution to the detection task, and dynamically assigns weighted coefficients. This method enhances the model’s detection robustness in complex backgrounds and effectively addresses the issue of insufficient emphasis on positive samples. Furthermore, through the optimization of the Path Aggregation Network (PAN) and the Bidirectional Feature Pyramid Network (BiFPN), efficient multi-scale feature fusion is achieved, further strengthening the model’s capacity to represent crack features at various scales. Experimental results indicate that the proposed GSB-YOLO model improves the mean average precision (mAP) in road crack detection tasks by 3.2%, demonstrating its significant application value in road crack detection and traffic safety assurance.https://doi.org/10.1038/s41598-025-11717-0Road crack detectionGSB-YOLOSMC2f moduleMulti-scale feature fusionYOLOv8n
spellingShingle Yuhao Wang
Heran Zhu
Yirong Wang
Jianping Liu
Jun Xie
Bi Zhao
Siyue Zhao
GSBYOLO: A lightweight Multi-Scale fusion network for road crack detection in complex environments
Scientific Reports
Road crack detection
GSB-YOLO
SMC2f module
Multi-scale feature fusion
YOLOv8n
title GSBYOLO: A lightweight Multi-Scale fusion network for road crack detection in complex environments
title_full GSBYOLO: A lightweight Multi-Scale fusion network for road crack detection in complex environments
title_fullStr GSBYOLO: A lightweight Multi-Scale fusion network for road crack detection in complex environments
title_full_unstemmed GSBYOLO: A lightweight Multi-Scale fusion network for road crack detection in complex environments
title_short GSBYOLO: A lightweight Multi-Scale fusion network for road crack detection in complex environments
title_sort gsbyolo a lightweight multi scale fusion network for road crack detection in complex environments
topic Road crack detection
GSB-YOLO
SMC2f module
Multi-scale feature fusion
YOLOv8n
url https://doi.org/10.1038/s41598-025-11717-0
work_keys_str_mv AT yuhaowang gsbyoloalightweightmultiscalefusionnetworkforroadcrackdetectionincomplexenvironments
AT heranzhu gsbyoloalightweightmultiscalefusionnetworkforroadcrackdetectionincomplexenvironments
AT yirongwang gsbyoloalightweightmultiscalefusionnetworkforroadcrackdetectionincomplexenvironments
AT jianpingliu gsbyoloalightweightmultiscalefusionnetworkforroadcrackdetectionincomplexenvironments
AT junxie gsbyoloalightweightmultiscalefusionnetworkforroadcrackdetectionincomplexenvironments
AT bizhao gsbyoloalightweightmultiscalefusionnetworkforroadcrackdetectionincomplexenvironments
AT siyuezhao gsbyoloalightweightmultiscalefusionnetworkforroadcrackdetectionincomplexenvironments