We propose a novel method for the detection of vibrationscaused by trains in an optical fiber buried nearby the railway track.Using optical time-domain reflectometry vibrations in the ground causedby different sources can be detected with high accuracy in time andspace. While several algorithms have been proposed in the literature fortrain tracking using OTDR signals they have not been tested on longerrecordings. The presented method learns the characteristic pattern in theFourier domain using a support vector machine (SVM) and it becomesmore robust to any kind of noise and artifacts in the signal. The point-based causal train tracking has two stages to minimize the influence offalse classifications of the vibration detection. Our technical contributionis the evaluation of the presented algorithm based on two hour longrecording and demonstration of open problems for commercial usage.


Papp, A., Wiesmeyr, C., Litzenberger, M., Garn, H., & Kropatsch, W. (2016). Train Detection and Tracking in Optical Time Domain Reflectometry (OTDR) Signals. Pattern Recognition, 320–331. https://doi.org/10.1007/978-3-319-45886-1_26