Real-time Vehicle Detection for Lane Change System of Autonomous Vehicle using 3D LiDAR in High-Definition Map

다국어 초록 (Multilingual Abstract)

Recently a lot of brand-new techniques of ADAS are being shown. Not only Longitudinal control but also lateral control system of the vehicle is getting commercialized by lots of motor company. This paper introduces a way of perception system of the autonomous vehicle which enables to change the lane. This system requires 3D LiDAR sensor and HD map (High-Definition map) for localization of the vehicle. It detects the objects on the road with 3D LiDAR sensor by making ROI (Region of interest) with curvilinear variable parameter acquired from the HD map. In addition, cluster the pointcloud and pick the vehicle using filters. Finally, this system acquires several key-data such as type and distance of the near-located-vehicle for the ego-vehicle to have enough data for lane change.