Inferring additional knowledge from QTCN relations
Information Sciences: an International Journal
A fuzzy spatio-temporal-based approach for activity recognition
ER'12 Proceedings of the 2012 international conference on Advances in Conceptual Modeling
MONET: modeling and querying moving objects in spatial networks
Proceedings of the Third ACM SIGSPATIAL International Workshop on GeoStreaming
Directional relations and frames of reference
Geoinformatica
| Hi-index | 0.00 |
In this paper, a basic set of motion relations capturing specific prototypical movements of vehicles on US road networks are introduced. Vehicle positional data collected from a geosensor network and stored in a spatio-temporal database serve as the basis for computing the relations that include isBehind, inFrontOf, driveBeside, and passBy. Relational SQL queries are used to derive the relations, returning information about pairs of moving objects and their relative positions. This information provides additional user contexts for binary vehicle patterns relative to a reference object. A framework for the kinds of moving objects that participate in these relations is supplied through an associated TransportationDevice ontology. Depending on the class of moving object, a relation such as isBehind captures scenarios that are facilitating or inhibiting with respect to the movement of traffic. For example, if a police car is known to be behind an automobile, the automobile typically slows to correspond with the legal speed limit. In this work, we show how linking the spatio-temporal database to an ontology can augment and extend the motion relation information, providing multi-granular perspectives of moving vehicles.