A Real-Time Similarity Measure Model for Multi-source Trajectories

Abstract

In order to solve the problem that the similarity of asynchronous multi-source multi-track cannot be measured effectively, a new trajectory similarity model for asynchronous multi-source multi-track is proposed in this paper. Based on the idea of searching the potential matched data points under the spatial and temporal constraints, the optimal matched point is determined from the sets of nearly matched points by setting a certain spatial threshold and temporal threshold, and the similarity measure between multi-source trajectories is obtained. The spatial and temporal relation between potential matched points from multi-source trajectories has been totally considered. The temporal slots between potential mapped points are allowed, reducing the complexity sharply and ensuring the trajectory mapping accuracy. The application to a real data set shows that the model can evaluate the similarity of multi-source trajectory effectively, and its time cost is lower than traditional methods.