Trajectories similarity measurements based on spatio-temporal patterns
Mesures de similarité de trajectoires basées sur l'utilisation de patrons spatio-temporels =
Résumé
Mobile objects are now equipped with sensors allowing real time monitoring of their movements. Nowadays, the data produced by these sensors can be stored in spatio-temporal databases. Data mining on these stored positions allows to infer the behaviour of these mobile objects (spatio-temporal patterns) and to analyze positions and trajectories of mobile objects. Using these patterns, unusual situations can be detected. Two kinds of unusual situations are distinguished: unusual positions and unusual trajectories. This article defines similarity index based on spatial and temporal measures using fuzzy logic.