SLAM (Simultaneous Localization and Mapping) technology has been studied to look for current position and generate map
simultaneously in computer vision research area. The technique is one of the most widely researched topics in Robotics and has
become applicable to UAV (unmanned aerial vehicle) recently for autonomous navigation flight. In case of indoor environment
which cannot support GPS information, SLAM is especially necessary to recognize current position. However it is difficult to
generate large-scale map because of cumulative error of the sensors. In this paper, we introduce large-scale 3D map generation
system in development for indoor autonomous navigation flight based on HectorSLAM which is available in ROS (Robot
Operating System).