Railway transportation plays an important role in the national economy of China. As a basic component of railway infrastructure, railway fasteners play a crucial role in connecting various components. However, traditional manual inspection methods cannot guarantee the accuracy and efficiency of fastener detection, which poses potential threats to railway safety. In recent years, the widespread application of deep learning technology has brought new opportunities for fastener detection. YOLOv8 algorithm is currently one of the popular object detection algorithms with excellent detection accuracy and speed. ShuffleNetV2 network structure is a lightweight convolutional neural network that can significantly reduce model parameters and computational complexity without sacrificing detection accuracy. Therefore, the purpose of this study is to improve the detection speed and accuracy of YOLOv8 algorithm by combining its lightweight feature with ShuffleNetV2, and to propose a lightweight fastener detection method based on ShuffleNetV2 and YOLOv8, and introduce SE attention mechanism to enhance feature extraction. The method was trained and tested using 2km data images collected from a certain section of Nanjing subway, and its effectiveness was verified through experiments. The results showed that the proposed method was 2.8% higher than the official YOLOv8 source code, and increased by 5.5% and 3.4% compared with YOLOv7 and YOLOv5, respectively. In addition, the proposed method has a smaller model and is more suitable for lightweight fastener detection on embedded devices. This research has important significance for improving the automation of fastener detection, reducing computational parameters, and improving detection speed.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    Lightweight Detection of Fasteners with YOLOv8 Combined with ShuffleNetV2


    Additional title:

    Lect. Notes Electrical Eng.


    Contributors:
    Yang, Jianwei (editor) / Yao, Dechen (editor) / Jia, Limin (editor) / Qin, Yong (editor) / Liu, Zhigang (editor) / Diao, Lijun (editor) / Duan, Jiaming (author) / Bai, Tangbo (author) / Lv, Baile (author) / Zong, Hao (author)

    Conference:

    International Conference on Electrical and Information Technologies for Rail Transportation ; 2023 ; Beijing, China October 19, 2023 - October 21, 2023



    Publication date :

    2024-01-09


    Size :

    10 pages





    Type of media :

    Article/Chapter (Book)


    Type of material :

    Electronic Resource


    Language :

    English




    EDGS-YOLOv8: An Improved YOLOv8 Lightweight UAV Detection Model

    Min Huang / Wenkai Mi / Yuming Wang | DOAJ | 2024

    Free access

    Night traffic flow detection method based on ShuffleNetv2-YOLOv5s

    CHEN DONG / CHEN MAO | European Patent Office | 2024

    Free access

    Lightweight YOLOv8 Networks for Driver Profile Face Drowsiness Detection

    Zhang, Meng / Zhang, Fumin | Springer Verlag | 2024


    Lightweight YOLOv8 Networks for Driver Profile Face Drowsiness Detection

    Zhang, Meng / Zhang, Fumin | Springer Verlag | 2024