Different advanced driver assistance systems (ADAS) are available in today's serial vehicles and provide important contribution in comfort-oriented as well as safety-oriented functions. In 1999 the first ACC (adaptive cruise control) system was launched on serial passenger cars. Since then the variety of ADAS was raising steadily. Many systems have shown great effort in highway application such as Side Assist, Lane Keeping Assist, etc.
Nowadays research activities are addressing assistance for inner-city scenarios as well. A robust environment perception with sensors like radar, vision and laser is required, which is able to cope with the unstructured environment in urban scenarios. These and other demands require alternative and innovative sensor fusion concepts. Compared to Kalman filter approaches of object tracking grid-based techniques are of high potential for these scenarios. The idea of using the advantages of the compact description in object-based tracking and the advantages of grid-based environment description to yield to a hybrid object-/ grid-based approach is the focus of this research. One disadvantage of the classic grid-based fusion is the capability of moving objects. Hence a main challenge at this point is to develop an extended grid-based fusion which is responsive to dynamic information but still maintain the robustness against noise. This prepares a decisive basic for effective combination with the object-based method. The goal of this hybrid approach is to provide a better recognition of other traffic participants, drivable areas as well as static obstacles. This is a basic for predictive and effective action concepts to enable more safety and comfort for the driver.