Situation analysis for automotive pre-crash systems

According to the "World Report on Road Traffic Injury Prevention" jointly issued by the World Health Organization and the World Bank about 1.2 million people are killed and up to 50 million people are injured in road traffic accidents worldwide each year. While passive safety systems like the airbag are already deployed successfully to reduce fatalities and injuries, active safety systems assist the driver by issuing a warning or by taking corrective actions to either avoid a collision completely or, if impossible, to mitigate collision consequences. Today's radar sensors have the ability to detect and track objects with a high accuracy in range and velocity, therefore a collision warning system may consist of a radar sensor, a data processing unit and a model to describe possible evasion maneuvers. This allows to analyze the probability of a collision and to calculate the danger potential of the current situation. In this paper, such a system is proposed and it is verified with synthetic as well as real sensor data.