Blind Area Traffic Prediction Using High Definition Maps and LiDAR for Safe Driving Assist

Abstract

This paper proposes a collision prediction method based on blind area traffic prediction using high definition maps and 3D LiDAR(Light Detection and Ranging). To keep safe driving, it is necessary to predict risks and to ready to avoid it. There are so many intersections with low visibility in residential area. In such environments, crossing collision accidents are often occurred. The proposed method in this paper predicts the blind area traffic using lane network information and particle filter, and updates the predicted results using visibility information of 3D LiDAR. Finally, this paper illustrates experimental results in urban environments and the calculated safe speed using proposed method is compared with expert driver data.