Sensor fusion for cognitive automobiles

  • project group: Kognitive Automobile
  • funding:

    DFG SFB / TR 28

The goal of this project is to combine data from all available sensors in the test car (stereo-camera, infrared camera, DGPS, IMU, two LIDAR sensors and different weather sensors) to create a three dimensional, grid-based map of the surrounding, and to detect, track and classify relevant objects in this map. As a further step, the object hypotheses of other cooperating cognitive mobiles shall be communicated and fused with the self-obtained object hypotheses. To this end, it will be necessary to define and evaluate confidence measures in the hypotheses of all involved cognitive automobiles. To compare different methods quantifiably, a database of relevant traffic situations under varying weather conditions shall be created as a test set.