Map Matching Based on Trajectory Route Feature

Abstract

A new segment-wise map matching (MM) method for off-line application context is presented in this paper. Specifically, we suggest slicing each vehicle trajectory into segments by tracing multiple trajectory route features. Further for each segment, a global MM strategy is adopted to find the path that best matches the corresponding trajectory. Four existing MM algorithms are used to benchmark the performance of the proposed method, and the MM quality is measured by two metrics. They are (1) the accuracy ratio of successfully matched trajectory points against manually verified base trajectories and (2) the average MM speed. The experiment results conducted on a trajectory dataset containing 234 kilometers of GPS traces validated that the proposed algorithm has good matching efficiency and consistently outperforms the other benchmarks in terms of matching accuracy.