Information fusion based autonomous vehicle
- Research on core technologies of autonomous driving based on information convergence

 Period: 2014.05. ~ 2016.06. / Budget: 420 k$


  ● Iterative closest point (ICP) based map matching algorithm for precise localization


  ● Track to track fusion with camera and radars
        - Four short range radars (50m, 140 degree)
        - One long range radar (150m, 28 degree)
        - Camera (30 m, 65 degree)


  ● Traffic light recognition based on HD map and precise localization
        - High precision (no false positives), Slope conditions (uphill, downhill)
        - Results
           - Video clip1: day, complex background, uneven road
           - Video clip2: day, uphill, under the bridge


  ● Environment model generation for planning
        - Information retrieval from map and transmission
           - Road information: local path, lane, stop line, …
           - Event information: intersection, split region, …