The Train Sliding On-line Monitoring System Based On Image Recognition

In recent years, as the economy in China is developing rapidly, the strategic position of the railway traffic system becomes more and more important. Parking safety is the foundation and prerequisite to the development of railway transport. Today, rail transportation is developing to large-scale, high speed and automation directions, then, how to realize the remote on-line monitoring to prevent the significant loss that caused by the train brake failure becomes one of the issues need to be resolved in protecting the safety of the railway sector.This paper bases on the train sliding online system as the main research line. First, the train intelligence check wheels device online monitoring system structures which is based on ZigBee have been described. Then, the short of the plan and the reason why to give up eventually are analyzed in detail. At last, the paper introduces two on-line monitoring systems about train's sliding that based on Computer Image Sciences.The first kind of monitoring method is based on the three times spline curve track identification algorithms. This paper discusses something about the system structure,working principle,the application of the algorithm,the model of tracks,hardware,software design and the effect of image processing. Also the paper uses the form of contrasting to describe the realization of monitoring systems in some same section. The second approach is based on the corner detection. In this way it introduces the design principle and how to extract the angular point. The Canny operator's principle and realization,hardware structure,software design,the effect of image processing and the simulation results are introduced in detail.In the fifth chapter of the article, the advantages and disadvantages of the two methods are analyzed and compared, and the process of system is debugged. Then the on-line monitoring center of user-end visual interface is described, the implementation and results are also told. At the end of the article, I summarize the insufficiency of the system design and tell what need to improve.