Asynchronous track-to-track association algorithm based on dynamic time warping distance

In the distributed multi-target tracking system, the local sensors often begin working at different time and provide tracks at different rates with different communication delays. As a result, the local tracks from different sensors are usually asynchronous. The current solution is time registration before track association which leads to track synchronization. However, when synchronizing, the estimation error increases. This affects performance of track-to-track association. In this paper, tracks are treated as time series, and using dynamic time warping method (DTW) measures the distance between any two tracks. DTW is a much more robust distance measure for time series, allowing similar shapes to match even if they are out of phase in the time axis. Considering track-to-track association problems, and confining the search area of DTW when optimizing, a fast algorithm is obtained. This is a post-processing technique of tracks. In order to make track-to-track association more accurately after obtaining the track data from sensors, the algorithm is proposed so that track fusion can be implemented next. Simulation results show that the presented method can effectively solve the asynchronous track-to-track association problem.