Lane-level localization based on situational awareness on highway

This paper presents a lane-level localization algorithm based on situational awareness fusing radar, vision, GPS, and digital map on highway. Under the assumption that the maximum number of lanes is known by both GPS and digital map, the position of a vehicle is determined in a term of lane-level position via situational awareness on highway. The situational awareness aims at recognizing maneuver of the vehicle, guardrail, and surrounding vehicles. By fusion radar and vision sensors, it is proposed that the lane-level position can be determined inferring all outputs of situational awareness. The proposed algorithm is validated experimentally via field test data.