Datafusion of two driver assistance system sensors

This contribution deals with data fusion between two sensors for driver assistance; a lidar-based ACC-sensor and a CMOS-based vision sensor system for LDW. The main properties and experimental results of the proposed approach will be described. The first fusion task is to supply the ACC sensor with lane information, obtained by the vision system, to improve the relevant target determination and control strategy. Furthermore, the LDW-sensor is concurrently able to use and verify ACC target hypotheses by vision-based object detection and tracking. This goes along with the improvement of the estimation of lateral target-position and -dynamics. Several test drives figure out the capability of this multiple sensor system. The main focus lies on information processing in the vision sensor, i.e. lane and object detection. In addition, the fusion method and data association inside the object detection module is specified.