Vehicle-Mounted Rail Patrol Inspection Equipment Using a Combination of Line-Scan Cameras and Lasers
High-speed railways must maintain a high level of smoothness, stability, and reliability because these factors are directly linked to the safe operation of high-speed trains and the protection of people’s lives and property. However, it is challenging to accurately inspect such defects using a vision system due to variations in illumination and the fast speed of the trains. To acquire precise rail surface information during high-speed operations, we propose vehicle-mounted rail patrol inspection equipment (VPIE) based on a combination of line-scan cameras and lasers. We provide detailed introductions to the composition module and working mode of the equipment. The curvature of the standard ball in the line-scan image is utilized as the correction ratio for performing tension correction on the rail image. A novel Zernike moment subpixel edge extraction algorithm based on one-dimensional maximum entropy is put forth as a means of discerning the edge of the standard ball. The splicing and cutting methods of line-scan images in both horizontal and vertical directions are described in detail to obtain regular rail images. A local dodging detection algorithm based on the illumination nonuniformity coefficient is proposed to automatically judge the illumination uniformity of the image and screen the images that require dodging processing. This improves the visual effect and retains as much detail in the image as possible. The experimental results demonstrate that (1) the line-scan image correction accuracy is high, and the rail infrastructure in the image is effectively restored, with 93% of the correction error being less than one pixel, (2) the stitching accuracy of the line-scan image is high, and there is no obvious stitching trace; the stitching error in both directions is less than one pixel, and (3) the rail images processed by the algorithm are characterized by a high degree of clarity, brightness, and the absence of shadows and noise. Furthermore, the rail details remain well-preserved. The VPIE method enables the acquisition of rail images that meet the submillimeter accuracy required for the measurement of defects given high-speed motion conditions.
Authors
- Qingzhou Mao (ORCID: https://orcid.org/0000-0002-7948-2828)
- Tingli Fan
- Yingcai Zhang (ORCID: https://orcid.org/0000-0002-3086-2440)
- Xu Zhang (ORCID: https://orcid.org/0000-0002-2104-9845)
Institutions
- Wuhan University (CN)
Publication Details
- Journal
- Journal of Surveying Engineering
- Published
- 2024-12-11
- DOI
- https://doi.org/10.1061/jsued2.sueng-1522
- Citations
- 3
- Primary Topic
- Advanced Measurement and Detection Methods
- Type
- article
- Field-Weighted Citation Impact
- 0.63