Discovering underpasses from walking trajectories

With widely applying of the GPS-enabled electronic devices, people can obtain customized services from LBS providers. Meanwhile, their trajectory data is also gathered from day to day. Mining trajectory data to find some interesting knowledge becomes hot in recent years. In this paper, we study how to discover the underpasses in a city by analyzing the walking trajectory database. As the trajectory through the underpass is quite different from that on the ground, we devise a novel solution to deal with this issue. To the best of our knowledge, our work is the first piece of work in this topic. Experiments upon real-life data sets illustrate the effectiveness and efficiency of the proposed method.