Year: 2025
Type: article
Abstract: This study addresses the challenges of low efficiency in manual railway fastener inspection and susceptibility to lighting interference in image-based detection by proposing a line structured light-based 3D reconstruction and fault detection method, enabling precise identification of fastener anomalies and loosening. Key research contributions include: 3D Point Cloud Reconstruction Developed an improved region-growing algorithm for segmenting fastener component point clouds, constructing high-precision 3D models.For Type I resilient fasteners, a host computer software was implemented for scanning and reconstruction. Experimental results demonstrated 40% reduction in point cloud noise and 35% improvement in model completeness. Anomaly Detection Algorithm Proposed an ensemble classifier model to detect six categories of anomalies, including nut loss and resilient fastener missing.At a detection speed of 40 km/h, the system achieved 95% detection accuracy with only 3% false alarm rate. Loosening Detection Method Utilized resilient fastener registration algorithms to calculate gap distances and established a conversion relationship between nut loosening values and fastener displacement.Field tests confirmed <1 mm measurement error and RMS error of 0.32 mm, meeting high-speed railway maintenance standards.The developed detection system has been validated through experiments, showing 8× efficiency improvement over manual inspection, providing critical technical support for intelligent railway maintenance.
DOI: 10.71465/fias.22
Authors: Lun Sun, Binglin Li, Hanwu Shi, Yunlei Liu
Institution: Wuhan Railway Vocational College of Technology
Language: English
Subfield: Civil and Structural Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W4409591219