Options
1991
Book Article
Titel
Situationsgraphen zur Modellierung von Verkehrsituationen
Alternative
Situation graphs for modelling of traffic situations
Abstract
Autonomous vehicles and copilot systems which are to operate in the highly structured domain of urban road traffic need a variety of sensorial information about their environment. An important means to acquire that information is image sequence analysis. In this article, we propose situation graphs to organize the generic conceptual models of such an image sequence analysis system. Situation graphs explicitly represent all the traffic situations which the system has to recognize in order to achieve its task and they state what actions the system has to execute in these situations. The mapping of situation graphs to sensorial data,i.e. image sequences, is done by graph traversing and guides the interpretation process to select from the diversity of available information those aspects which are relevant to the actions which have to be executed in the current situation.