Automatic generation of intersection models from digital maps for vision-based driving on inner city intersections

Driver support in inner city road traffic still presents a considerable challenge for machine vision. Model-based machine vision becomes attractive in this context since it allows one to exploit the knowledge provided by a model in order to select relevant image structures. It requires, however, to make suitable models available to the computer vision system. This contribution shows that a digital road map for a commercial car navigation system can also be exploited, in order to generate road and intersection models which can be used by model-based tracking of related lane boundary markings in video image sequences recorded from within a driving vehicle. Results obtained by an implementation of this approach are presented.