Practice And Application Of Distance Metric In The Trajectory Clustering

Because of the large amount of trajectory data produced by mobile devices, there is an increasing need for mechanisms to extract knowledge from this data. Most existing works have focused on the geometric properties of trajectories, but recently emerged the concept of semantic trajectories, in which the background geographic information is integrated to trajectory sample points. In this new concept, trajectories are observed as a set of stops and moves, where stops are the most important parts of the trajectory. Based on the need for cluster analysis, this paper studies the trajectory similarity measure practice and application method.First, the temporal database-related knowledge, including spatial and temporal data modeling, temporal database query method, the index build overview, GIS Spatial Data GIS technology as well as an overview of the spatio-temporal database concepts and moving objects on the road clustering method, and finally with step density-based DBscan algorithm as an example of the trajectory of the initial clustering, and the impact of DBscan algorithm for clustering effect of the input parameters.Then, the paper proposes a spatial distance algorithm movement locus T-IMHD, hausdorff principle based algorithm, using any point on a track to another track from the minimum track neutrons constituting a subset minimum distance space using noising threshold method, and finally to the experimental data for validation.Finally, a place classification algorithm based on Bayesian network, definitions and concepts moving object trajectory stays district, and proposed merging algorithm to identify multi-stacking area, and then introduces the concept of stacking area to identify points and using Bayesian network classification categories for stays semantic similarity identify areas to determine the trajectory.