Similarity measurement of moving object trajectories

Abstract

To study the similarity between moving object trajectories is important in many applications, e.g., to find the clusters of moving objects which share the same moving pattern, and infer the future locations of a moving object from its similar trajectories. To define the similarity between moving objects is a challenging task, since not only their locations change but also their speed and semantic features vary. In this paper, we propose a novel approach to measure the similarity between trajectories. The similarity is defined based on both geographic and semantic features of movements. Our approach can be used to detect trajectory clusters and infer future locations of moving objects.