A method for real-time monitoring of running trains using Ф-OTDR and the improved Canny.

A novel recognition method for optical fiber vibration signals is proposed to monitor the running trains real-time and accurately using the phase-sensitive optical-time-domain reflectometer (Ф-OTDR) and the improved Canny. Firstly, all kinds of noise is reduced by the cubical smoothing algorithm with five-point approximation to enhance timefrequency characteristics of vibration signals and shorten the calculation time when compared with the stationary wavelet transform. Then it extracts vibration signals of the running trains by differencing and thresholding. Finally, we introduce the improved Canny edge detection to make the edges clearer. Through on-site test of SHUO HUANG Railway, the error of this method is less than 10m, which can be used as a secondary measure to monitor the running trains. [ABSTRACT FROM AUTHOR]