A new roadmap approach to automatic path planning for mobile robot navigation

This paper describes a new road map approach to find a collision-free path for an AGV (automatic guided vehicle), taking into account the smoothness, length and available space restrictions. The proposed approach uses a top view image of a room in which there are some obstacles to be avoided. First of all the obstacles in the image are thickened using morphological techniques, to obtain the free space of the room. Secondly, this space is skeletonized to obtain a graph representation of AGV workspace, enriched with geometrical information of the free space. Finally, the best path is chosen using an optimization technique with a new criteria based on length, availability of free space, curvature and smoothness.<<ETX>>