Građevinar, Vol. 77 No. 01., 2025.
Original scientific paper
https://doi.org/10.14256/JCE.3925.2023
Non-contact intelligent detection technology for railway arch bridge performance degradation based on UAV Image recognition
Shifu Wang
Shaopeng Yang
Qi Wang
Lingfeng Luo
Feng Wang
Abstract
Bridges are crucial components of high-speed railway projects, and their structural integrity significantly impacts the operational safety of high-speed railways. This paper introduces a non-contact intelligent detection technology for assessing the deterioration of high-speed railway bridges using unmanned aerial vehicle (UAV) image recognition. The methodology involves collecting image data using a UAV and digital camera and processing them technically to generate consistent point-cloud data. Subsequently, these data are integrated into a unified point-cloud model through point-cloud alignment. Finally, a refined three-dimensional (3D) model of a high-speed railway bridge was developed by fusing heterogeneous data through live 3D reconstruction. The method has the advantages of high detection speed and fewer personnel requirements; this technology can be used for daily monitoring of the technical basis and can arrange a small number of personnel to complete the daily inspection. The empirical results demonstrate that this inspection method is not constrained by skylight points and provides a real-time and highly efficient reflection of the conditions of the bridge. The recognition accuracy and image acquisition range satisfy the inspection requirements for the operation and maintenance of high-speed railway bridges.
Keywords
high-speed railway bridge; bridge faults; non-contact measurement; UAV
Hrčak ID:
329021
URI
Publication date:
18.2.2025.
Visits: 267 *