Online Map-Matching Algorithm Using Object Motion Laws

Modern city intelligent transportation system urgently demands high accuracy real-time map matching methods. Because of inaccurately measured locations, a GPS point can be assigned to many road segments depending on the topology of the road network and GPS measurement, especially in the complex urban traffic environment. Taking some object motion laws, such as speed limitations and acceleration constraints, into consideration can significantly reduce the matching errors. However, most current map matching approaches didn't pay attention to these laws, leading to inefficiency and inaccuracy. This paper proposes a novel map matching algorithm, called Object Motion Laws Map Matching(OMLMM). The object motion laws integrate various moving status of vehicles including direction, velocity, and acceleration, which are the major characteristics for the matching process. With these effective laws, the OMLMM could efficiently solve the map matching problem in a strongly constrained urban traffic environment. We evaluate the accuracy and running time of our algorithm using ground-truth data. The experiment results show that the object motion laws constrained map matching algorithm outperforms existing methods regarding both accuracy and output delay.