Detection method for leaky cable clamps in high-speed railway tunnels based on improved YOLOv5
ZHANG Yunzuo
ZHANG Luqi
SUN Yuchuan
LI Yingxu
WANG Ning
Abstract：To address the low efficiency of manual inspection and the difficulty of processing massive data when detecting leaky cable clamps in high-speed railway tunnels,this study proposes a detection method based on an improved YOLOv5 model.First,a Ghost Spatial Convolutional Spatial Pyramid(GSCSP)module is designed,where Ghost convolution replaces standard convolution to reduce fea-ture redundancy and achieve model lightweighting.Second,a lightweight Efficient Channel Attention Network(ECANet)is integrated to strengthen the model's ability to distinguish clamp features from com-plex tunnel backgrounds and improve the detection accuracy of small objects.Then,a structured chan-nel pruning strategy is applied,in which redundant channels are pruned based on the scaling factors of Batch Normalization(BN)layers,achieving a lightweight architecture while maintaining model accu-racy.Finally,a dataset covering various states of leaky cable clamps under real high-speed railway op-eration scenarios is constructed.Data diversity is enhanced through methods such as Gaussian noise ad-dition and data stitching,providing a richer and more robust sample foundation for subsequent model training.Experimental results show that the improved model reduces the number of parameters by 72.1%while preserving detection accuracy and real-time performance.The findings provide a refer-ence for the intelligent operation and maintenance of railway communication equipment.
Keywords：computer applicationclamp detectionchannel attentionmodel pruning
Publication Date：2025-12-30
Online Publishing Date：2026-02-02(First online date of this platform, not the publication date of the document)
Pages：9( 55-63 )