Accurate and complete rail extraction from mobile laser scanning (MLS) data is currently a fundamental and challenging problem for its application on the railway. By using the track knowledge, a signed cylindrical neighborhood difference is defined as the rail descriptor and then proposed a new rail extraction algorithm from MLS data. It can extract accurate, continuous, and complete railhead, which is most critical for the rail geometric parameter and centerline, of the entire railway. Moreover, it can successfully extract the railhead of the main-line, including the curve section with different superelevation, and turnout. A 3-km long trunk railway, including main-line and turnout, straight line and curve line, located in the southwest of China is selected to test the performance of the proposed rail extraction algorithm. The experimental results show that the proposed algorithm can correctly extract the railhead of the whole railway, with an overall accuracy (F-measure) of 88.73%. Its accuracy is improved by 42.68% compared with the rail extraction algorithm based on spherical neighborhood difference.
With the rapid development of high-speed railway, there are many problems with the traditional railway slab assessment method. The traditional method is slow, and its precision is limited by the precision of specified tools for railway slab inspection. Scholars have developed a variety of inspection systems for railway slab geometry. Since those systems’ precision assessment relies on railway slab testing tools that are complex for operation, this paper proposes a novel method to assess the precision of an intelligent slab inspection system itself by using the spatial position deviation between the point cloud of a benchmark slab and the corresponding digital 3D model. The proposed method takes the RMSE of the deviation value of points in the key surfaces as the evaluation index. The key surfaces are the two shoulder surfaces and the rail-bearing surface of the rail-bearing platform, which can be extracted by the region growing algorithm associated with surface normals. Based on the real point cloud processed by an intelligent slab inspection system, the experimental results show that the system can align the slab point cloud to its corresponding 3D digital model. The deviation is distributed on the model uniformly, and its precision is 0.1 mm. In addition, this procedure is consistent with that of general slab inspection and can be used as a self-verification tool for daily precision evaluation of the system itself.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.