Designing Huge Repositories of Moving Vehicles Trajectories for Efficient Extraction of Semantic Data

The rapid development of digital cameras equippedwith video analytics software is providing the availability of largeamount of traffic data describing the trajectories traced by eachvehicle and person within a scene. These data offer enormouspotential when coupled with a querying system able to extractsynthetic but meaningful information as those obtained by spatiotemporalqueries; the latter allow, for instance, to select all thosetrajectories passing through some parts of the scene, even ingiven sequences, and adding restrictions on the properties of theobjects (the category of the vehicles, their color and size, and soon).In this paper we propose a novel system for efficiently storingand querying large amounts of 3D data (trajectories over time),specifically designed for making possible the formulation of awide variety of spatio-temporal 3D queries. The method is basedon a novel 3D data schema which is reconducted to a set of 2Dschemata, being the latter the only ones available in currentlyready-to-use database environments.An implementation of the system over PostGIS is presentedin this paper, together with a performance assessment on ahuge trajectory database. The obtained results confirm theeffectiveness of the proposed approach and its applicability toreal applications.