Enhanced environmental perception by inter-vehicle data exchange

Abstract

Inter-vehicle data communication promises a multitude of improvements in next generation driver assistance systems. The field of view of one car can be significantly extended by information from the environmental sensors of other cars in its vicinity. However, many of the algorithmic problems are not yet satisfactorily solved, or have not even been properly addressed in research. This paper deals with various aspects of this topic. We propose to exchange two types of information: First, the state and positioning information of each of the participating vehicles and second, the objects detected and tracked by each of the vehicle's data processing systems. Object data is acquired by a radar sensor and tracked by Kalman filtering. Each participating vehicle incorporates all remote information in a global description of the environment. This description is based upon a digital map. For this purpose, a reliable map matching has been developed. Results for real traffic scenes are presented.