Key technologies for intelligent and safer cars - From motion estimation to predictive collision avoidance

One of the research areas that has received more and more interest during the last years is the development of driver assistant systems and semi-autonomous cars. However, densely populated environments like city centers are still a challenge for the operation of such systems. In this paper, we present approaches to two of the major tasks for autonomous driving in urban environments: self-localization and egomotion estimation and detection of dynamic objects such as cars and pedestrians. For each of these tasks we present a summary of the techniques we employ and results on real data. All modules have been implemented and tested on our autonomous car platform SmartTer.