Tracking and classifying objects with DAS data along railway

Abstract

Using distributed acoustic sensing data from a day of field testing on a fiber-optic cable along a railroad track in Norway, we detect and track cars and trains moving along a segment of the cable where the road runs parallel to the railroad tracks. We develop a method for automatic detection of events using signal processing, thresholding and density-based clustering, and then put data picks into a Kalman filter variant known as joint probabilistic data association filter for multiple object tracking and classification. Statistical model parameters are specified using in-situ labeling data along with the fiber-optic signals. Running the algorithm over time, we automatically track about 100 cars and 20 trains per hour. The velocities of cars coming from a zone with higher speed limit tend to be larger (35 km/h) than that of cars going in the opposite direction (30 km/h).