RoadGraph: High level sensor data fusion between objects and street network

The RoadGraph is a graph based environmental model for driver assistance systems. It integrates information from different sources like digital maps, onboard sensors and V2X communication into one single model about the vehicle´s environment. At the moment of information aggregation some function independent situation analysis is done. In this paper we look at techniques for lane-precise map-matching even with moderate GPS reception using distinct information sources. We also analyze the concepts of aggregating objects from different sources with a-priori knowledge of the street-layout. Results of this novelty approach are shown.