Radar depth association with vision detected vehicles on a highway

No single sensor can independently predict the 3-D environment around us. A vision sensor helps locate objects in a 2-D plane. However, estimating distance using one vision sensor has a limitation. A single radar sensor returns the range of objects accurately; however, the complexity and cost increase if good spatial resolution is required. Thus a radar does not indicate which range corresponds to which object. In this paper, we associate vision detected objects to radar returned echoes, focusing on highways, estimating the 3-D locations of cars around an ego vehicle. This information would help cars driving autonomously to maneuver around.