Pedestrian classification with 24 GHz chirp sequence radar

Driver assistance systems based on radar are available today in series production in consumer vehicles. The radar sensor based automotive applications already help to reduce road accidents but is usually limited to highway scenarios. However, there are too many fatalities in public roads. Most of them involve pedestrians and happen in urban road situations. For assisting drivers in those cases, a novel 24 GHz Radar Sensor has been developed within the EU-funded project ARTRAC. The main focus of this paper is to present the pedestrian classification scheme that is incorporated within this sensor. Additionally, the derived requirements and the role of the classification process within the cognitive collision mitigation system is pointed out.