Hybrid Approach for a Future Environmental Representation for Advanced Driver Assistance Systems

In recent years of research activities within the field of Advanced Driver Assistance Systems (ADAS), it became more and more obvious, that the present standard representation of the current state of a vehicle's environment is not sufficient anymore. Experience has shown that for functions addressed in the years ahead, there is a strong need for more dense environmental models, than the commonly used object lists. Many proposals have been made towards this direction; these are discussed in this work. Based on this overview, we propose a so called “Hybrid Environmental Model” approach for meeting requirements of future ADAS functions by complementing the existing solutions with a dense, grid-based representation. Additionally, methods are discussed which enhance the efficiency of data transport and therefore enable a potential series application of this approach.