Polarimetric millimeter wave imaging radar and traffic scene interpretation

Improving comfort and safety is one of the major concerns of car manufacturers. Efforts have been made to realize prototypes of driver assistance, intelligent cruise control and collision avoidance systems. Those systems must be able to detect obstacles with a high reliability even under bad weather conditions and poor visibility. Decreasing prices for millimeter wave components especially for monolithic millimeter wave integrated circuits make radars operating in those frequency ranges a very interesting sensor solution for the various system demands. Due to the short wavelength, even real aperture imaging with high lateral resolution becomes possible with compact systems. This was the starting point for the development of a prototypic automotive radar imaging sensor. Based upon stationary measurements with a fully polarimetric instrumentation radar, the specific imaging features in road traffic environment have been investigated. The usually unintentional multi-path effects caused by ground reflections can be exploited profitably. Echoes from the vehicle floors give information about their longitudinal dimensions. The various applications of automotive radar sensors like autonomous intelligent cruise control and collision avoidance require an automatic interpretation of the perceived data. Important processing tasks in this context like vehicle classification and roadside detection can be performed and are validated with measured data. Outlooks of the described techniques concerning sensor data fusion and the combination of sensor data with navigation information are discussed.