Fusion of data from the object-detecting sensors of an autonomous vehicle

This paper describes Kalman filter based approaches for the fusion of data from object detecting sensors for an autonomous vehicle. The vehicle sensor system for object detection consists of several dissimilar sensors observing the whole environment of the vehicle. These sensors work independently from each other and are equipped with their own target recognition system such that pre-filtered target data is transmitted to the sensor fusion system. Three different approaches for the fusion of sensor data have been investigated and are described and evaluated in this paper.