The Train Sliding On-line Monitoring System Of Anti-interference Based On Image Processing

As the rapid development of economy and society, the railway transport system plays a more and more important role in passenger transport and the transport of cargo in China. Train parking security is the guarantee of the development of the railway transportation, Scotch block is the main equipment of the train brake, due to the Scotch block falling off automatically or man-made theft and other reasons, it is familiar that the train happtens to slip, which often causes a serious accident. In order to realize the state of remote monitor of the train, the research of the train sliding on-line monitoring system of railway based on image processing to guarantee transportation system securely.This paper based on train sliding on-line monitoring system of anti-disturbance as the main line. First, the paper introduces two monitoring methods:based on the train tracks identification method based on sliding monitoring and corner detection monitoring method of sliding train. In the two monitoring method based on improving the system robustness. The train stations is the dynamic changes of the environment, such as the sun illuminate Angle changes will cause changes of light, the light of the change is very easy to create the vicinity of on-line monitoring system; In a train passed or high winds, the hard to avoid cameras have jitter, a shaking of the camera can cause monitoring image dithering, monitoring system to cause disturbance, thus cause the train sliding on-line monitoring system miscalculation.Considering the influence of the two main interference, this paper puts forward block background update methods and image stabilization. Background updates block matching algorithm based on the block to judge whether continuous steady state, if it is not stable, eliminate the light and other outside interference. To dither algorithm is in DaoSanJiao method and the background detection based on image stabilization of improvement, makes the trains sliding monitoring system better ruled out dithering on the monitoring of interference improve the system robustness.