基于GNSS/INS与激光雷达的缓存池建图

为了解决智能车、无人机、自主机器人对外界环境感知的先验信息不足的问题。本文提出了一种使用惯性导航(INS)与卫星导航(GNSS)辅助多线激光雷达传感器对环境进行自主三维重建的算法。该算法首先采用一种基于GNSS\\INS的线性差值法与四元数插值法并行算法将激光雷达点云转换到唯一坐标系下,纠正点云随时间的漂移;通过基于缓存池的建图方法对周围环境进行点云分割与拼接,对得到的子地图进行体素滤波,最终得到丰富环境信息的三维重建地图。通过机器人操作系统(ROS)在移动平台上进行验证,实验结果表明:该方法成功的对外界环境进行自主三维建图,适用于智能化载体对未知环境下的感知。

Cache pool mapping based on GNSS/INS/LiDAR

In order to solve the problem of the insufficient information of the smart car,the drone,and the autonomous robot aware of the external environment is insufficient. This paper presents an algorithm for autonomous 3D reconstruction of the environment using Inertial Navigation System (INS) and Global Navigation Satellite System(GNSS) assisted multi-line LiDAR sensors.The algorithm firstly uses a linear difference method based on GNSS\\INS and a parallel algorithm of quaternion interpolation to convert the LiDAR point cloud to a unique coordinate system to correct the drift of the point cloud with time;The method performs point cloud segmentation and splicing on the surrounding environment,and performs voxel filtering on the obtained sub-map,and finally obtains a three-dimensional reconstruction map enriching environmental information. The robot operating system (ROS) is verified on the mobile platform. The experimental results show that the method successfully performs independent 3D mapping on the external environment,which is suitable for the perception of the intelligent carrier in the unknown environment.