Camera and lidar fusion for road intersection detection

The road intersection is crucial for local path planning and position control for autonomous vehicle in urban environments. Camera and lidar fusion method is proposed for road intersection detections in this paper. First, the lane information from lidar and three monocular cameras are extracted. Second, the virtual predict trajectory space for autonomous vehicle is constructed, and the lane information is projected in this space. Finally, the possible intersection branches are detected from the fusion of lane information. The algorithm presented in this paper is proved to be both robust and fast by experiments.