Semi-automated Map Generation Using Space Segmentation and Generalized Voronoi Graph for Robot Navigation

In this paper, we propose a semi-automatic method for creating maps and creating destination nodes, which are essential for the driving of mobile robots. First, the process converts a grid map that can be recognized by the mobile robot based on the prior drawing for the new environment and divides the movable space from the grid map by applying a space segmentation approach. Next, through generalized Voronoi graph, the mobile robot creates a target node that can be reached without collision on the map, matches the node to belong to the divided spatial information, and then layer the destination node. It is tested that space segmentation and node creation were automated from the actual CAD drawings, and it was confirmed that the destination node information where the mobile robot can move can be initially set through this process.