Three-Dimensional Cooperative Mapping for Connected and Automated Vehicles

Abstract

To improve the efficiency of three-dimensional (3-D) LIDAR mapping, multiple robots cooperation has been considered in mapping of large areas. In order to merge the local maps acquired by independent robots, it is crucial to identify common areas among the local maps. In this article, we propose a novel 3-D cooperative mapping approach for connected and automated vehicles that use only 3-D LIDAR data to cooperatively create globally consistent maps among multiple robots. Assuming each individual vehicle is able to create its own local map, the cooperative mapping process is divided into two main tasks: The common area detection task and the map merging task. We first significantly reduce the time cost of common area detection task by utilizing ground information as features. Further, the accuracy of map merging is guaranteed by whole frame matching of the common area frame pairs generated in the previous step. The proposed method is tested and validated by experiments in real-world environment for various scenarios; results show that the efficiency and accuracy of mapping are both improved compared to the individual robot mapping approach.