Global Map Generation using LiDAR and Stereo Camera for Initial Positioning of Mobile Robot

Abstract

A lot of researches have been made on how to know the position of the mobile robot when it knows the initial position of the mobile robot. For the robot to be completely unmanned, the robot also needs to find out its own initial position. To do this, it is necessary to obtain enough data to estimate the initial position without previous data. In Simultaneous Localization and Mapping (SLAM), Light Detection and Ranging (LiDAR) is often used to obtain accurate map for mobile robot. In this case, it is difficult to find the initial position because there is little information at initial start-up. On the other hand, stereo camera has the advantage that it can acquire more information than LiDAR by acquiring spatial information with the same principle as the human eye. However, the obtained 3D spatial information has a disadvantage of low precision. The purpose of this paper is to form a 3D map to be used for finding the initial position by linking LRF information to compensate for the low accuracy of the 3D map made only by the stereo camera in the environment.