2008
EMSS 2008 Proceeding
Self-learning navigation maps based upon data-driven models using recorded heterogeneous GPS tracks
Authors: Clemens Novak, Barbara Franz, Herwig Mayr, Michal Vesely
Abstract
We present our innovative approach to keep navigationmaps up to date by deducing map changes from recordedGPS tracks using adequate models and rules.First, we describe, how models for receiver, mobilityand terrain can be generated from adequately preprocessedrecorded GPS tracks. These models are usedby a server in order to predict plausible extensions ofavailable navigation maps. In order to allow for multimodaltrack sources (pedestrians, automobilists, bicyclists,horseback riders, etc.), geometrical matches have tobe further checked for plausibility. We give examples ofsuch plausibility rules we have developed for this purpose.The main benefits of our development are bettermaps and better guidance for various classes of possibleusers, from pedestrian, over cross-country skier, to busdriver, to name just a few.
I3M Scientific Sponsors
I3M Industrial Sponsors
I3M Media Sponsors
